Real-time Top-view People Counting Based on A Kinect and NVIDIA Jetson TK1 Integrated Platform

Size: px
Start display at page:

Download "Real-time Top-view People Counting Based on A Kinect and NVIDIA Jetson TK1 Integrated Platform"

Transcription

1 2016 IEEE 16th International Conference on Data Mining Workshops Real-time Top-view People Counting Based on A Kinect and NVIDIA Jetson TK1 Integrated Platform Guangqin Li, Peng Ren, Xinrong Lyu, and He Zhang College of Information and Control Engineering China University of Petroleum (East China) Qingdao, China ligq@s.upc.edu.cn, pengren@upc.edu.cn, lvxr@upc.edu.cn, hezhang@s.upc.edu.cn Abstract In this paper, we describe how to establish an embedded framework for real-time top-view people counting. The development of our system consists of two parts, i.e. establishing an embedded signal processing platform and designing a people counting algorithm for the embedded system. For the hardware platform construction, we use Kinect as the camera and exploit NVIDIA Jetson TK1 board as the embedded processing platform. We describe how to build a channel to make Kinect for windows version 2.0 communicate with Jetson TK1. Based on the embedded system, we adapt a water filling based scheme for top-view people counting, which integrates head detection based on water drop, people tracking and counting. Gaussian Mixture Model is used to construct and update the background model. The moving people in each video frame are extracted using background subtraction method. Additionally, the water filling algorithm is used to segment head area as Region Of Interest(ROI). Tracking and counting people are performed by calculating the distance of ROI center point before and after the frame. The whole framework is flexible and practical for real-time application. Keywords-Kinect; Jetson TK1; Top-view; People counting; Waterfilling. I. INTRODUCTION People counting is an important task for marketing research [1] (e.g. pedestrian traffic management [2], tourists flow estimation and people flow monitoring [3] and security applications [4], [5] (in evacuation, estimating people flow is essential to save lives). In the literature, most high precision algorithms of people counting are performed on computers [6]. However, in some scenarios such as on a bus, the space for placing equipments are limited. Therefore, how to develop real-time people counting frameworks on a small-sized embedded device is highly demanded. In the past few years, novel algorithms were rarely implemented via embedded platforms. It was because that the previously popular embedded development platforms [7], [8] have limited resources. Most of these embedded platforms could not meet the real-time requirements. Therefore, the implementation of sophisticated counting algorithms on these embedded platforms are technically difficult. Nowadays, with the recent development of embedded technologies, there are many embedded computing platforms which can be performed in parallel operation. Raspberry Pi, BeagleBone Black and Jetson TK1 are currently popular embedded development platforms on the market, which are widely used and more powerful. Especially Jetson TK1 is armed with rich interfaces and powerful computing capabilities. In this paper, we set up a top-view camera for recording walking people because occlusion is naturally avoided. Most of popular stereo cameras [9], [10] are mainly using Kinect for windows version 1.0, Kinect for windows version 2.0 (hereinafter referred to as Kinect V2), Primesense and etc. Kinect V2 is a latest camera, which can collect higher resolution image and range accuracy. Since the heads are always closer to the camera than other parts of the body in top view of Kinect sensor [11], detecting heads is posed as a problem of finding the suitable local minimum regions in the depth image. The depth image [1] is shown in Fig. 1. When some people walk in the counting area, they are very close to each other and partial occlusion might occur. In this situation, depth images play a more important role in people detection and extraction than color images because color features are rarely capable of identifying occlusions. In the literature, counting by detection is the most popular approach. Once people have been detected, the counting problem is solved by numbering the detected objects. However, object detection also has two unsolved problems. Firstly, one object may correspond to different bounding boxes and currently the commonly used NMS (non-maximum suppression) [12] for deciding a bound box tends to become invalid in complex situation where many people have interactions with each other. Secondly, some objects are too difficult to be detected, because they may be partially or even wholly occluded. Both problems affect the counting performance. Our framework adapts water filling method which is motivated by the water filling process [12]. It is a good solution to the above two problems arising in detection. In the water filling process, the water moves away from the heave and out to the nearby hollow under the force of gravity until the gravitational potential energy cannot be reduced any more. We simulate the rain by generating the raindrop according to an uniform distribution. Once /16 $ IEEE DOI /ICDMW

2 Range of Depth 8.0m Range of Detection 4.5m 255 Kinect 2.0 Jetson TK1 Camera 0.5m 0 Pixel Value Set Up Communication Platform Figure 1. Kinect view range (left) and example depth image (right) a raindrop arrives, we compare its landing spot with its neighborhood and find the descent direction until it cannot descend any more, then the number of raindrop at the balance spot increases. Water filling algorithm can detect the top of the head more accurate. II. PROPOSED SYSTEM In this section we introduce our proposed system for people-counting. First, a communication channel between camera and computing platform is built to obtain the depth image. Second, the depth image processing algorithm is developed to achieve people counting. The overview of proposed system is shown in Fig. 2. A. Hardware platform structures Our video capture device is a Microsoft Kinect for Windows version 2.0. Kinect V2 is an upgraded version of Kinect for windows version 1.0, which is integrated with depth sensors, color sensors and infrared sensors. Compared to the prevision version, Kinect V2 has a lot of improvements. For example, the color image resolution is promoted from to , the depth of the image resolution is promoted from to , and the detection range is more accurate. Our computing platform is NVIDIA s production Jetson TK1. Jetson TK1 runs on NVIDIA s latest Tegra TK1 SoC (system on chip). It has four 32-bit ARM Cortex- A15 general purpose cores running at 2.3GHz, a low-power Cortex-A15 shadow core that runs when the system is not under heavy load, and 192 Kepler GPU cores running at 852MHz for graphical horsepower. In addition to the powerful image processing capabilities of Tegra TK1, we employ the undersampling strategy and propose a statistical tracking algorithm for real-time image processing. Jetson TK1 is capable of performing GPU computation via NVIDIA s CUDA (Compute Unified Device Architecture) computing platform. Jetson TK1 is a powerful platform for computer vision and CUDA applications and may have potential applications in robotics, medicine, security, automotive, and even national defense. On the other hand, the configuration for operating kinect V2 is as follows: Windows 8.1 or higher Figure 2. Acquire Images Background Model Water Filling Head Detection People Tracking And Counting The proposed people counting system Microsoft Visual Studio 2013 Community Edition with Update 2 or higher Microsoft.NET Framework 4.5 [13] Kinect for Windows SDK 2.0 [13] USB 3.0 Though both Jetson TK1 and Kinect V2 have their own merit for real-time data collection and processing, they are not compatible to each other. To establish an embedded platform integrating Kinect V2 and Jetson TK1, we build a communication channel between them. The details for build the communication channel is described as follows: Step 1): Initialization video capture device on PC. In order to use the Kinect V2 to collect data for Jetson TK1, we first initialize Kinect V2 by a Windows 8.1 PC. This should be done before physically connecting the Kinect V2 to the Jetson TK1 for the first time. We download the Kinect drivers and the Kinect for Windows SDK 2.0 [13], 469

3 Figure 3. Download toolkits Install tools Compile Sample Micro USB Micro USB Internet Jetson TK1 Flash OS Install tools Test compiled sample Jetson TK1 configurations via communications with a PC from the Kinect official website and install them according to the installation guide. We then connect kinect V2 with PC through USB 3.0 and running an example in the installation directory. If some images can be captured by the Kinect sensor and illustrated on the PC screen, the initialization is successfully. Step 2): Systematic configuration of the Jetson TK1. Jetson TK1 is an embedded platform with the operating system Linux Ubuntu on board. However, the raw Linux system have limited ability of maneuvering the hardware resources on board for data processing. To enhance the data processing power of the platform, we need to perform a systematic configuration of the board, and install some toolkits that are used for manipulating GPUs and developing signal processing algorithms, etc. NVIDIA corporation provide a setup package for configuring Jetson TK1, which is Jetson TK1 Development Pack. Jetson TK1 Development Pack includes host (Ubuntu Desktop) and target (Jetson TK1) development tools, APIs and packages (OS images, tools, CUDA, OpenCV, APIs, middleware, samples, documentation) for developing with NVIDIA Tegra K1 on the NVIDIA Jetson TK1 platform. The whole configuration process requires an iterative interplay between a PC with Linux Ubuntu System and the NVIDIA Jetson TK1 through an USB 3.0 connection. Additionally, both of them require an internet access. The overview of this process is shown in Fig. 3. Once we finish this step, we disconnect the Jetson TK1 from the PC. The Jetson TK1 will then operate as an individual platform and will not be connected to PC again in our task. Step 3): Acquire depth and color images. We connect the Kinect V2 to the Jetson TK1 via USB 3.0. Before we exploit it as our embedded platform for processing the data captured by the Kinect V2, we need to install some other toolkits, such as Libfreenect2, Libusb and TurboJPEG, to Jetson TK1. Libfreenect2 is a crossing platform driver of Kinect V2. Libfreenect2 provide a lot of APIs that can be used to control the camera. Libusb is a C library that gives applications easy access to USB devices in many different operating systems. TurboJPEG APIs wraps libjpeg-turbo and provides an easyto-use interface for compressing and decompressing JPEG images in memory. These toolkits are downloaded directly to the Jetson TK1 via the on board Linux Ubuntu The installation of these toolkits enable the Jetson TK1 to obtain color and depth video smoothly from the Kinect camera. The integration of Kinect V2 and Jetson TK1 thus obtained forms an infrastructural platform which can potentially implements image/video processing algorithms in real-time on board. The Kinect V2 is possible to capture high resolution RGB and depth images/videos, and the GPUs in Jetson TK1 are capable of conducting parallel computing for algorithms, which in turn possibly result in real-time outputs. B. People counting scheme In this subsection, we introduce the people counting method, which is mainly divided into three parts. 1) Build the background model using Gaussian Mixture Model [10]. 2) Detect people head using water filling method [12]. 3) Apply people tracking and counting. 1) Background model and foreground extraction: Based on the infrastructural platform established in the previous subsection, we start developing people counting algorithms. We commence by developing a motion detection strategy for segmenting the moving objects from depth images using background subtraction technique. The change of the illumination or motion has strong influence on the intensity of color image but not on the depth information. Therefore, we use Gaussian Mixture Model (GMM) [10] for background modeling which can be achieved by computing its probability as follows: P (z t )= L ω i,t η(z t,μ i,t, ) (1) i,t i=1 where z t represent one pixel in the original image containing Z pixels. ω i,t and L is the weighted Gaussian and the number of Gaussian, respectively. η is the probability of Gaussian function defined by: η (z t,μ, ) = 1 (2π) Z e 1 2 (zt μ t ) T 1 (zt μ t ) Once background is constructed, whereafter foreground will be obtained. In the foreground region, there may be a foreground object overlapping, then head detection is used to separate the overlapping targets. (2) 470

4 2) Head detection: According to [12], we exploit the water filling algorithms for head detection. Compared to alternative detection algorithms (e.g. Mean Shift and Random Forest), water filling has more advantages in antiinterference and detection of children. we represent the depth value of pixel (x, y) byf(x, y). Due to the noise of depth image, f(x, y) can be discontinuous or even null-valued. This method presents a measure function g(x, y) to measure f(x, y). g(x, y) is defined if and only if ε>0, which is an error controlling parameter (x i,y i ), (x j,y j ) subjected to (x i x j ) 2 +(y i y j ) <ε,iff(x i,y i ) f(x j,y j ): f(x i,y i )+g(x i,y i ) f(x j,y j )+g(x j,y j ) g(x i,y i ) g(x j,y j ) g(x i,y i ) 0, g(x j,y j ) 0 The function g(x, y) has three advantages: Robustness: Handle noises in the original depth data. Locality: Detect people in different heights. Scale-Invariance: Detect people head at any scales. The form of function f(x, y) can be seen as a land with humps and hollows. The raindrop in the hump will flow directly to the neighborhood hollow under force of gravity. Step by step, the hollow region will gather a lot of raindrops. The function g(x, y) reflects the quantity of raindrop at (x, y). After the rain stops, the regions with a lot of rain drop can be classified as a hollow. Water filling algorithm can be seen as the simulation of the above process. Furthermore, the water filling approach solve the basic problem that the quantity of raindrops is a continuous value and different raindrops may reach the land simultaneously, which lead to the difficulty of simulating their interactions. Therefore, water filling algorithm is an effective strategy for head detection. Supposed that there is a foreground object contour image patch with the size of M N, its depth function is f(x, y) and initialize the measure function as g(x, y) =0.LetT be the threshold, K be the number of raindrops. Provided (x m,y n ) is one of the neighbor of (x, y). Then we search the potential human head pixel (ˆx, ˆy) by minimizing the following score function O: min O(x m,y n )=f(x m,y n )+g(x m,y n ) f(x, y)+g(x, y) (3) which enables to obtain the minimum value at the point (ˆx, ˆy). IfO(x m,y n ) is greater than zero, g(x, y) increases one and the loop is end, otherwise take point (x m,y n ) as a new point (x, y) and perform the above operations again. After all the K raindrops are set to their stable places, we obtain the final measure function g(x, y). Then thresholding and contour analysis operations are applied to g(x, y). By setting threshold T on g(x, y), contour analysis is employed to find contours corresponding to each people. Finally, take A B B A Figure 4. A 1, A 2, B 1 and B 2 demonstrate the center of each rectangular for two person in two successive frames, d 11, d 12, d 21 and d 22 represent the distance between A 1, A 2, B 1 and B 2, respectively. all the contours extracted as the input of people tracking and counting. 3) People tracking and counting: In this subsection, a simple and effective method was proposed to track the head position and count people. In addition, people s walking direction (i.e. in or out) can be judged by the tracking scenario. Fig. 4 illustrate the overview of the proposed people tracking and counting process. We first find the center of each depth image patch for different person. For instance, A i and A j demonstrate the two different persons in the previous frame. While B i and B j represent the two head patches in current depth image. We then calculate the pairwise distance of head patches above as follows: d ij = (x i x j ) 2 +(y i y j ) 2 (4) where the x { } and y { } exhibit the coordinates of the center associated with corresponding depth patch. By calculating the distance, it is obviously that the same depth patch of one person possess the less distance among the previous frame and the current frame than that with another. Fig. 6 illustrate the threshold selection in our calculating approach. Data B represents the distance of one person among the current and previous frame, while Data A is the distance between different people on the same frame. Thus, it is obviously obtain the threshold which leads to a high performance of distinction for different people. By setting a suitable distance threshold, when the d ij distance is smaller than the threshold, A i is the same path of B j in current depth image. Only if the tracking scheme continued for more than twelve frames, the counter increases one for one person. If the distance is larger than the distance threshold value, we refer to the rectangle of the depth patch as a new people. III. EXPERIMENTAL RESULTS We conducted three groups of comparative experiments in order to compare with other algorithms and platforms. The 471

5 Figure 5. Proposed system detection results Figure 7. DSP system detection results Figure 6. Threshold selection first one is the comparison of different hardware platforms. The second one is the comparison of different algorithms. The third one is to test the accuracy of proposed system for real-time stream of people counting. A. Experimental comparisons on alternative infrastructural platforms The work presented in [8] was conducted on a DSP embedded platform for people stream statistics. They proposed a people counting method for estimating the number of people crowd in large spaces such as lobbies, wholesaleand super-stores. The their main algorithm flow was as follows: 1) Capturing frame from camera; 2) Constructing background model; 3) Updating background model; 4) Extracting foreground; 5) Constructing the weighting model; 6) Accumulating the weighed pixels and counting people. In order to facilitate comparison, we remove the tracking part in our experiments. We perform this set of experiments on our Jetson TK1 which is equipped with 32-bit ARM Cortex-A15 (2.3GHz), 192 Kepler GPU (852MHz) and 16GB memory. The results are shown in Fig. 5 and Fig. 7. In the experimental results, it is shown that our results are in better agreement with the real statistics. Especially, for the first fifty frames, the DSP embedded platform has a bigger error in people statistics. B. Experimental comparisons based on alternative algorithms This set of experiments are performed on a PC machine with Intel Xeon E5 CPU (3.5GHz) and 64GB memory for comparing algorithm performances. The water filling algorithm is compared with the a commonly used people counting scheme presented in [10] based on depth and color images, whose procedure is as follows: 1) Capturing depth and color image from Kinect; 2) Motion detection blob analysis; 3) People segmentation use U-disparity and color matching; 4) People tracking and counting. The experimental results are shown in Table 1. According to Table 1, it is obvious that the accuracy of the water filling algorithm is significantly higher than the algorithm proposed in [10]. Two groups of experiments were run in the same environment, which ruled out other factors except the algorithm. C. Experiments on proposed system This set of experiments are performed on the Jetson TK1 in a practically captured depth sequences. The first 300 frames of the image are used for training the background model. After this training procedure, the people counting starts operating. The people counting experimental results are shown in Fig. 8. The resulted statistics of proposed system on two different databases are demonstrated in Table 2. IV. CONCLUSION In this paper, we have constructed an embedded people counting system which physically integrates Kinect V2 and 472

6 Figure 8. People detection and counting results Jetson TK1 and algorithmically exploits a water filling scheme. We have presented a communication strategy for transferring data from Jetson TK1 to Kinect V2. We have adapted a water filling algorithm for human head detection and proposed an effective tracking and counting scheme. Experimental comparisons performed on various platforms with alternative algorithms have validated the superiority of our proposed system. Through different experimental setups, it has been shown that the proposed system has great potential applications in various practical situations. REFERENCES [1] T.-H. Chen, T.-Y. Chen, and Z.-X. Chen, An intelligent people-flow counting method for passing through a gate, in 2006 IEEE Conference on Robotics, Automation and Mechatronics. IEEE, 2006, pp [2] C. Hong, J. Yu, D. Tao, and M. Wang, Image-based threedimensional human pose recovery by multiview localitysensitive sparse retrieval, IEEE Transactions on Industrial Electronics, vol. 62, no. 6, pp , [3] C. Hong, J. Yu, J. Wan, D. Tao, and M. Wang, Multimodal deep autoencoder for human pose recovery, IEEE Transactions on Image Processing, vol. 24, no. 12, pp , [4] N. Bernini, L. Bombini, M. Buzzoni, P. Cerri, and P. Grisleri, An embedded system for counting passengers in public transportation vehicles, in Mechatronic and Embedded Systems and Applications (MESA), 2014 IEEE/ASME 10th International Conference on. IEEE, 2014, pp [5] A. B. Chan, Z.-S. J. Liang, and N. Vasconcelos, Privacy preserving crowd monitoring: Counting people without people models or tracking, in Computer Vision and Pattern Recognition, CVPR IEEE Conference on. IEEE, 2008, pp [6] X. Zhao, E. Delleandrea, and L. Chen, A people counting system based on face detection and tracking in a video, in Advanced Video and Signal Based Surveillance, AVSS 09. Sixth IEEE International Conference on. IEEE, 2009, pp [7] A. G. Vicente, I. B. Munoz, P. J. Molina, and J. L. L. Galilea, Embedded vision modules for tracking and counting people, IEEE Transactions on Instrumentation and Measurement, vol. 58, no. 9, pp , [8] H.-C. Chen, Y.-C. Chang, N.-J. Li, C.-F. Weng, and W.-J. Wang, Real-time people counting method with surveillance cameras implemented on embedded system, in Proceedings of the World Congress on Engineering and Computer Science, vol. 1, [9] B. Alessandro, D. S. Luigi, and A. Pietro, People tracking using a time of flight depth sensor, in 2006 IEEE International Conference on Video and Signal Based Surveillance. IEEE, 2006, pp [10] C. Wateosot, N. Suvonvorn et al., top-view based people counting using mixture of depth and color information., in The second asian conference on information systems, ACIS. Citeseer, [11] T. Van Oosterhout, S. Bakkes, and B. J. Kröse, Head detection in stereo data for people counting and segmentation. in VISAPP, 2011, pp [12] X. Zhang, J. Yan, S. Feng, Z. Lei, D. Yi, and S. Z. Li, Water filling: Unsupervised people counting via vertical kinect sensor, in Advanced Video and Signal-Based Surveillance (AVSS), 2012 IEEE Ninth International Conference on. IEEE, 2012, pp [13] A. Jana, Kinect for windows SDK programming guide. Packt Publishing Ltd,

Water Filling: Unsupervised People Counting via Vertical Kinect Sensor

Water Filling: Unsupervised People Counting via Vertical Kinect Sensor 2012 IEEE Ninth International Conference on Advanced Video and Signal-Based Surveillance Water Filling: Unsupervised People Counting via Vertical Kinect Sensor Xucong Zhang 1 Junjie Yan 1 Shikun Feng 1

More information

People Counting based on Kinect Depth Data

People Counting based on Kinect Depth Data Rabah Iguernaissi, Djamal Merad and Pierre Drap Aix-Marseille University, LSIS - UMR CNRS 7296, 163 Avenue of Luminy, 13288 Cedex 9, Marseille, France Keywords: Abstract: People Counting, Depth Data, Intelligent

More information

TR An Overview of NVIDIA Tegra K1 Architecture. Ang Li, Radu Serban, Dan Negrut

TR An Overview of NVIDIA Tegra K1 Architecture. Ang Li, Radu Serban, Dan Negrut TR-2014-17 An Overview of NVIDIA Tegra K1 Architecture Ang Li, Radu Serban, Dan Negrut November 20, 2014 Abstract This paperwork gives an overview of NVIDIA s Jetson TK1 Development Kit and its Tegra K1

More information

Robust color segmentation algorithms in illumination variation conditions

Robust color segmentation algorithms in illumination variation conditions 286 CHINESE OPTICS LETTERS / Vol. 8, No. / March 10, 2010 Robust color segmentation algorithms in illumination variation conditions Jinhui Lan ( ) and Kai Shen ( Department of Measurement and Control Technologies,

More information

An Approach for Real Time Moving Object Extraction based on Edge Region Determination

An Approach for Real Time Moving Object Extraction based on Edge Region Determination An Approach for Real Time Moving Object Extraction based on Edge Region Determination Sabrina Hoque Tuli Department of Computer Science and Engineering, Chittagong University of Engineering and Technology,

More information

Automatic Shadow Removal by Illuminance in HSV Color Space

Automatic Shadow Removal by Illuminance in HSV Color Space Computer Science and Information Technology 3(3): 70-75, 2015 DOI: 10.13189/csit.2015.030303 http://www.hrpub.org Automatic Shadow Removal by Illuminance in HSV Color Space Wenbo Huang 1, KyoungYeon Kim

More information

THE LEADER IN VISUAL COMPUTING

THE LEADER IN VISUAL COMPUTING MOBILE EMBEDDED THE LEADER IN VISUAL COMPUTING 2 TAKING OUR VISION TO REALITY HPC DESIGN and VISUALIZATION AUTO GAMING 3 BEST DEVELOPER EXPERIENCE Tools for Fast Development Debug and Performance Tuning

More information

Moving Object Detection and Tracking for Video Survelliance

Moving Object Detection and Tracking for Video Survelliance Moving Object Detection and Tracking for Video Survelliance Ms Jyoti J. Jadhav 1 E&TC Department, Dr.D.Y.Patil College of Engineering, Pune University, Ambi-Pune E-mail- Jyotijadhav48@gmail.com, Contact

More information

FAST HUMAN DETECTION USING TEMPLATE MATCHING FOR GRADIENT IMAGES AND ASC DESCRIPTORS BASED ON SUBTRACTION STEREO

FAST HUMAN DETECTION USING TEMPLATE MATCHING FOR GRADIENT IMAGES AND ASC DESCRIPTORS BASED ON SUBTRACTION STEREO FAST HUMAN DETECTION USING TEMPLATE MATCHING FOR GRADIENT IMAGES AND ASC DESCRIPTORS BASED ON SUBTRACTION STEREO Makoto Arie, Masatoshi Shibata, Kenji Terabayashi, Alessandro Moro and Kazunori Umeda Course

More information

An Adaptive Threshold LBP Algorithm for Face Recognition

An Adaptive Threshold LBP Algorithm for Face Recognition An Adaptive Threshold LBP Algorithm for Face Recognition Xiaoping Jiang 1, Chuyu Guo 1,*, Hua Zhang 1, and Chenghua Li 1 1 College of Electronics and Information Engineering, Hubei Key Laboratory of Intelligent

More information

PEOPLE IN SEATS COUNTING VIA SEAT DETECTION FOR MEETING SURVEILLANCE

PEOPLE IN SEATS COUNTING VIA SEAT DETECTION FOR MEETING SURVEILLANCE PEOPLE IN SEATS COUNTING VIA SEAT DETECTION FOR MEETING SURVEILLANCE Hongyu Liang, Jinchen Wu, and Kaiqi Huang National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Science

More information

Reliable Human Detection and Tracking in Top-View Depth Images

Reliable Human Detection and Tracking in Top-View Depth Images 2013 IEEE Conference on Computer Vision and Pattern Recognition Workshops Reliable Human Detection and Tracking in Top-View Depth Images Michael Rauter Austrian Institute of Technology Donau-City-Strasse

More information

Connected Component Analysis and Change Detection for Images

Connected Component Analysis and Change Detection for Images Connected Component Analysis and Change Detection for Images Prasad S.Halgaonkar Department of Computer Engg, MITCOE Pune University, India Abstract Detection of the region of change in images of a particular

More information

A Street Scene Surveillance System for Moving Object Detection, Tracking and Classification

A Street Scene Surveillance System for Moving Object Detection, Tracking and Classification A Street Scene Surveillance System for Moving Object Detection, Tracking and Classification Huei-Yung Lin * and Juang-Yu Wei Department of Electrical Engineering National Chung Cheng University Chia-Yi

More information

Learning the Three Factors of a Non-overlapping Multi-camera Network Topology

Learning the Three Factors of a Non-overlapping Multi-camera Network Topology Learning the Three Factors of a Non-overlapping Multi-camera Network Topology Xiaotang Chen, Kaiqi Huang, and Tieniu Tan National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy

More information

Adaptive Background Mixture Models for Real-Time Tracking

Adaptive Background Mixture Models for Real-Time Tracking Adaptive Background Mixture Models for Real-Time Tracking Chris Stauffer and W.E.L Grimson CVPR 1998 Brendan Morris http://www.ee.unlv.edu/~b1morris/ecg782/ 2 Motivation Video monitoring and surveillance

More information

Human Motion Detection and Tracking for Video Surveillance

Human Motion Detection and Tracking for Video Surveillance Human Motion Detection and Tracking for Video Surveillance Prithviraj Banerjee and Somnath Sengupta Department of Electronics and Electrical Communication Engineering Indian Institute of Technology, Kharagpur,

More information

Research on the Application of Digital Images Based on the Computer Graphics. Jing Li 1, Bin Hu 2

Research on the Application of Digital Images Based on the Computer Graphics. Jing Li 1, Bin Hu 2 Applied Mechanics and Materials Online: 2014-05-23 ISSN: 1662-7482, Vols. 556-562, pp 4998-5002 doi:10.4028/www.scientific.net/amm.556-562.4998 2014 Trans Tech Publications, Switzerland Research on the

More information

Accurate 3D Face and Body Modeling from a Single Fixed Kinect

Accurate 3D Face and Body Modeling from a Single Fixed Kinect Accurate 3D Face and Body Modeling from a Single Fixed Kinect Ruizhe Wang*, Matthias Hernandez*, Jongmoo Choi, Gérard Medioni Computer Vision Lab, IRIS University of Southern California Abstract In this

More information

Figure 1 shows unstructured data when plotted on the co-ordinate axis

Figure 1 shows unstructured data when plotted on the co-ordinate axis 7th International Conference on Computational Intelligence, Communication Systems and Networks (CICSyN) Key Frame Extraction and Foreground Modelling Using K-Means Clustering Azra Nasreen Kaushik Roy Kunal

More information

Research on Evaluation Method of Video Stabilization

Research on Evaluation Method of Video Stabilization International Conference on Advanced Material Science and Environmental Engineering (AMSEE 216) Research on Evaluation Method of Video Stabilization Bin Chen, Jianjun Zhao and i Wang Weapon Science and

More information

Real-Time Human Detection using Relational Depth Similarity Features

Real-Time Human Detection using Relational Depth Similarity Features Real-Time Human Detection using Relational Depth Similarity Features Sho Ikemura, Hironobu Fujiyoshi Dept. of Computer Science, Chubu University. Matsumoto 1200, Kasugai, Aichi, 487-8501 Japan. si@vision.cs.chubu.ac.jp,

More information

Pedestrian counting in video sequences using optical flow clustering

Pedestrian counting in video sequences using optical flow clustering Pedestrian counting in video sequences using optical flow clustering SHIZUKA FUJISAWA, GO HASEGAWA, YOSHIAKI TANIGUCHI, HIROTAKA NAKANO Graduate School of Information Science and Technology Osaka University

More information

Vehicle Detection Method using Haar-like Feature on Real Time System

Vehicle Detection Method using Haar-like Feature on Real Time System Vehicle Detection Method using Haar-like Feature on Real Time System Sungji Han, Youngjoon Han and Hernsoo Hahn Abstract This paper presents a robust vehicle detection approach using Haar-like feature.

More information

Background subtraction in people detection framework for RGB-D cameras

Background subtraction in people detection framework for RGB-D cameras Background subtraction in people detection framework for RGB-D cameras Anh-Tuan Nghiem, Francois Bremond INRIA-Sophia Antipolis 2004 Route des Lucioles, 06902 Valbonne, France nghiemtuan@gmail.com, Francois.Bremond@inria.fr

More information

Tri-modal Human Body Segmentation

Tri-modal Human Body Segmentation Tri-modal Human Body Segmentation Master of Science Thesis Cristina Palmero Cantariño Advisor: Sergio Escalera Guerrero February 6, 2014 Outline 1 Introduction 2 Tri-modal dataset 3 Proposed baseline 4

More information

Evaluation of Moving Object Tracking Techniques for Video Surveillance Applications

Evaluation of Moving Object Tracking Techniques for Video Surveillance Applications International Journal of Current Engineering and Technology E-ISSN 2277 4106, P-ISSN 2347 5161 2015INPRESSCO, All Rights Reserved Available at http://inpressco.com/category/ijcet Research Article Evaluation

More information

Exploiting Depth Camera for 3D Spatial Relationship Interpretation

Exploiting Depth Camera for 3D Spatial Relationship Interpretation Exploiting Depth Camera for 3D Spatial Relationship Interpretation Jun Ye Kien A. Hua Data Systems Group, University of Central Florida Mar 1, 2013 Jun Ye and Kien A. Hua (UCF) 3D directional spatial relationships

More information

Design of a Dynamic Data-Driven System for Multispectral Video Processing

Design of a Dynamic Data-Driven System for Multispectral Video Processing Design of a Dynamic Data-Driven System for Multispectral Video Processing Shuvra S. Bhattacharyya University of Maryland at College Park ssb@umd.edu With contributions from H. Li, K. Sudusinghe, Y. Liu,

More information

Object Tracking Algorithm based on Combination of Edge and Color Information

Object Tracking Algorithm based on Combination of Edge and Color Information Object Tracking Algorithm based on Combination of Edge and Color Information 1 Hsiao-Chi Ho ( 賀孝淇 ), 2 Chiou-Shann Fuh ( 傅楸善 ), 3 Feng-Li Lian ( 連豊力 ) 1 Dept. of Electronic Engineering National Taiwan

More information

Human Detection. A state-of-the-art survey. Mohammad Dorgham. University of Hamburg

Human Detection. A state-of-the-art survey. Mohammad Dorgham. University of Hamburg Human Detection A state-of-the-art survey Mohammad Dorgham University of Hamburg Presentation outline Motivation Applications Overview of approaches (categorized) Approaches details References Motivation

More information

A Feature Point Matching Based Approach for Video Objects Segmentation

A Feature Point Matching Based Approach for Video Objects Segmentation A Feature Point Matching Based Approach for Video Objects Segmentation Yan Zhang, Zhong Zhou, Wei Wu State Key Laboratory of Virtual Reality Technology and Systems, Beijing, P.R. China School of Computer

More information

Classification of objects from Video Data (Group 30)

Classification of objects from Video Data (Group 30) Classification of objects from Video Data (Group 30) Sheallika Singh 12665 Vibhuti Mahajan 12792 Aahitagni Mukherjee 12001 M Arvind 12385 1 Motivation Video surveillance has been employed for a long time

More information

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc. Research on motion tracking and detection of computer vision ABSTRACT KEYWORDS

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc. Research on motion tracking and detection of computer vision ABSTRACT KEYWORDS [Type text] [Type text] [Type text] ISSN : 0974-7435 Volume 10 Issue 21 BioTechnology 2014 An Indian Journal FULL PAPER BTAIJ, 10(21), 2014 [12918-12922] Research on motion tracking and detection of computer

More information

Robot localization method based on visual features and their geometric relationship

Robot localization method based on visual features and their geometric relationship , pp.46-50 http://dx.doi.org/10.14257/astl.2015.85.11 Robot localization method based on visual features and their geometric relationship Sangyun Lee 1, Changkyung Eem 2, and Hyunki Hong 3 1 Department

More information

Research on Recognition and Classification of Moving Objects in Mixed Traffic Based on Video Detection

Research on Recognition and Classification of Moving Objects in Mixed Traffic Based on Video Detection Hu, Qu, Li and Wang 1 Research on Recognition and Classification of Moving Objects in Mixed Traffic Based on Video Detection Hongyu Hu (corresponding author) College of Transportation, Jilin University,

More information

International Journal of Computer Engineering and Applications, Volume XII, Issue I, Jan. 18, ISSN

International Journal of Computer Engineering and Applications, Volume XII, Issue I, Jan. 18,   ISSN International Journal of Computer Engineering and Applications, Volume XII, Issue I, Jan. 18, www.ijcea.com ISSN 2321-3469 SURVEY ON OBJECT TRACKING IN REAL TIME EMBEDDED SYSTEM USING IMAGE PROCESSING

More information

CSE/EE-576, Final Project

CSE/EE-576, Final Project 1 CSE/EE-576, Final Project Torso tracking Ke-Yu Chen Introduction Human 3D modeling and reconstruction from 2D sequences has been researcher s interests for years. Torso is the main part of the human

More information

Face Recognition Using Vector Quantization Histogram and Support Vector Machine Classifier Rong-sheng LI, Fei-fei LEE *, Yan YAN and Qiu CHEN

Face Recognition Using Vector Quantization Histogram and Support Vector Machine Classifier Rong-sheng LI, Fei-fei LEE *, Yan YAN and Qiu CHEN 2016 International Conference on Artificial Intelligence: Techniques and Applications (AITA 2016) ISBN: 978-1-60595-389-2 Face Recognition Using Vector Quantization Histogram and Support Vector Machine

More information

An efficient face recognition algorithm based on multi-kernel regularization learning

An efficient face recognition algorithm based on multi-kernel regularization learning Acta Technica 61, No. 4A/2016, 75 84 c 2017 Institute of Thermomechanics CAS, v.v.i. An efficient face recognition algorithm based on multi-kernel regularization learning Bi Rongrong 1 Abstract. A novel

More information

Large-scale Surveillance System based on Hybrid Cooperative Multi-Camera Tracking

Large-scale Surveillance System based on Hybrid Cooperative Multi-Camera Tracking Open Journal of Applied Sciences, 2013, 3, 79-84 Published Online March 2013 (http://www.scirp.org/journal/ojapps) Large-scale Surveillance System based on Hybrid Cooperative Multi-Camera Tracking Xiao

More information

A Novel Real-Time Feature Matching Scheme

A Novel Real-Time Feature Matching Scheme Sensors & Transducers, Vol. 165, Issue, February 01, pp. 17-11 Sensors & Transducers 01 by IFSA Publishing, S. L. http://www.sensorsportal.com A Novel Real-Time Feature Matching Scheme Ying Liu, * Hongbo

More information

ФУНДАМЕНТАЛЬНЫЕ НАУКИ. Информатика 9 ИНФОРМАТИКА MOTION DETECTION IN VIDEO STREAM BASED ON BACKGROUND SUBTRACTION AND TARGET TRACKING

ФУНДАМЕНТАЛЬНЫЕ НАУКИ. Информатика 9 ИНФОРМАТИКА MOTION DETECTION IN VIDEO STREAM BASED ON BACKGROUND SUBTRACTION AND TARGET TRACKING ФУНДАМЕНТАЛЬНЫЕ НАУКИ Информатика 9 ИНФОРМАТИКА UDC 6813 OTION DETECTION IN VIDEO STREA BASED ON BACKGROUND SUBTRACTION AND TARGET TRACKING R BOGUSH, S ALTSEV, N BROVKO, E IHAILOV (Polotsk State University

More information

Subpixel Corner Detection Using Spatial Moment 1)

Subpixel Corner Detection Using Spatial Moment 1) Vol.31, No.5 ACTA AUTOMATICA SINICA September, 25 Subpixel Corner Detection Using Spatial Moment 1) WANG She-Yang SONG Shen-Min QIANG Wen-Yi CHEN Xing-Lin (Department of Control Engineering, Harbin Institute

More information

MOVING OBJECT DETECTION USING BACKGROUND SUBTRACTION ALGORITHM USING SIMULINK

MOVING OBJECT DETECTION USING BACKGROUND SUBTRACTION ALGORITHM USING SIMULINK MOVING OBJECT DETECTION USING BACKGROUND SUBTRACTION ALGORITHM USING SIMULINK Mahamuni P. D 1, R. P. Patil 2, H.S. Thakar 3 1 PG Student, E & TC Department, SKNCOE, Vadgaon Bk, Pune, India 2 Asst. Professor,

More information

Guided Image Super-Resolution: A New Technique for Photogeometric Super-Resolution in Hybrid 3-D Range Imaging

Guided Image Super-Resolution: A New Technique for Photogeometric Super-Resolution in Hybrid 3-D Range Imaging Guided Image Super-Resolution: A New Technique for Photogeometric Super-Resolution in Hybrid 3-D Range Imaging Florin C. Ghesu 1, Thomas Köhler 1,2, Sven Haase 1, Joachim Hornegger 1,2 04.09.2014 1 Pattern

More information

Stereo Vision Based Traversable Region Detection for Mobile Robots Using U-V-Disparity

Stereo Vision Based Traversable Region Detection for Mobile Robots Using U-V-Disparity Stereo Vision Based Traversable Region Detection for Mobile Robots Using U-V-Disparity ZHU Xiaozhou, LU Huimin, Member, IEEE, YANG Xingrui, LI Yubo, ZHANG Hui College of Mechatronics and Automation, National

More information

3D HAND LOCALIZATION BY LOW COST WEBCAMS

3D HAND LOCALIZATION BY LOW COST WEBCAMS 3D HAND LOCALIZATION BY LOW COST WEBCAMS Cheng-Yuan Ko, Chung-Te Li, Chen-Han Chung, and Liang-Gee Chen DSP/IC Design Lab, Graduated Institute of Electronics Engineering National Taiwan University, Taiwan,

More information

High-Fidelity Augmented Reality Interactions Hrvoje Benko Researcher, MSR Redmond

High-Fidelity Augmented Reality Interactions Hrvoje Benko Researcher, MSR Redmond High-Fidelity Augmented Reality Interactions Hrvoje Benko Researcher, MSR Redmond New generation of interfaces Instead of interacting through indirect input devices (mice and keyboard), the user is interacting

More information

Detection and Classification of Vehicles

Detection and Classification of Vehicles Detection and Classification of Vehicles Gupte et al. 2002 Zeeshan Mohammad ECG 782 Dr. Brendan Morris. Introduction Previously, magnetic loop detectors were used to count vehicles passing over them. Advantages

More information

Moving Object Detection for Video Surveillance

Moving Object Detection for Video Surveillance International OPEN ACCESS Journal Of Modern Engineering Research (IJMER) Moving Object Detection for Video Surveillance Abhilash K.Sonara 1, Pinky J. Brahmbhatt 2 1 Student (ME-CSE), Electronics and Communication,

More information

A new Computer Vision Processor Chip Design for automotive ADAS CNN applications in 22nm FDSOI based on Cadence VP6 Technology

A new Computer Vision Processor Chip Design for automotive ADAS CNN applications in 22nm FDSOI based on Cadence VP6 Technology Dr.-Ing Jens Benndorf (DCT) Gregor Schewior (DCT) A new Computer Vision Processor Chip Design for automotive ADAS CNN applications in 22nm FDSOI based on Cadence VP6 Technology Tensilica Day 2017 16th

More information

Lecture 19: Depth Cameras. Visual Computing Systems CMU , Fall 2013

Lecture 19: Depth Cameras. Visual Computing Systems CMU , Fall 2013 Lecture 19: Depth Cameras Visual Computing Systems Continuing theme: computational photography Cameras capture light, then extensive processing produces the desired image Today: - Capturing scene depth

More information

A Hand Gesture Recognition Method Based on Multi-Feature Fusion and Template Matching

A Hand Gesture Recognition Method Based on Multi-Feature Fusion and Template Matching Available online at www.sciencedirect.com Procedia Engineering 9 (01) 1678 1684 01 International Workshop on Information and Electronics Engineering (IWIEE) A Hand Gesture Recognition Method Based on Multi-Feature

More information

TRANSPARENT OBJECT DETECTION USING REGIONS WITH CONVOLUTIONAL NEURAL NETWORK

TRANSPARENT OBJECT DETECTION USING REGIONS WITH CONVOLUTIONAL NEURAL NETWORK TRANSPARENT OBJECT DETECTION USING REGIONS WITH CONVOLUTIONAL NEURAL NETWORK 1 Po-Jen Lai ( 賴柏任 ), 2 Chiou-Shann Fuh ( 傅楸善 ) 1 Dept. of Electrical Engineering, National Taiwan University, Taiwan 2 Dept.

More information

Motion Detection Algorithm

Motion Detection Algorithm Volume 1, No. 12, February 2013 ISSN 2278-1080 The International Journal of Computer Science & Applications (TIJCSA) RESEARCH PAPER Available Online at http://www.journalofcomputerscience.com/ Motion Detection

More information

Object Tracking System Using Motion Detection and Sound Detection

Object Tracking System Using Motion Detection and Sound Detection Object Tracking System Using Motion Detection and Sound Detection Prashansha Jain Computer Science department Medicaps Institute of Technology and Management, Indore, MP, India Dr. C.S. Satsangi Head of

More information

Designing Applications that See Lecture 7: Object Recognition

Designing Applications that See Lecture 7: Object Recognition stanford hci group / cs377s Designing Applications that See Lecture 7: Object Recognition Dan Maynes-Aminzade 29 January 2008 Designing Applications that See http://cs377s.stanford.edu Reminders Pick up

More information

Traffic Signs Recognition using HP and HOG Descriptors Combined to MLP and SVM Classifiers

Traffic Signs Recognition using HP and HOG Descriptors Combined to MLP and SVM Classifiers Traffic Signs Recognition using HP and HOG Descriptors Combined to MLP and SVM Classifiers A. Salhi, B. Minaoui, M. Fakir, H. Chakib, H. Grimech Faculty of science and Technology Sultan Moulay Slimane

More information

DYNAMIC BACKGROUND SUBTRACTION BASED ON SPATIAL EXTENDED CENTER-SYMMETRIC LOCAL BINARY PATTERN. Gengjian Xue, Jun Sun, Li Song

DYNAMIC BACKGROUND SUBTRACTION BASED ON SPATIAL EXTENDED CENTER-SYMMETRIC LOCAL BINARY PATTERN. Gengjian Xue, Jun Sun, Li Song DYNAMIC BACKGROUND SUBTRACTION BASED ON SPATIAL EXTENDED CENTER-SYMMETRIC LOCAL BINARY PATTERN Gengjian Xue, Jun Sun, Li Song Institute of Image Communication and Information Processing, Shanghai Jiao

More information

Analysis Of Classification And Tracking In Vehicles Using Shape Based Features

Analysis Of Classification And Tracking In Vehicles Using Shape Based Features ISSN: 2278 0211 (Online) Analysis Of Classification And Tracking In Vehicles Using Shape Based Features Ravi Kumar Kota PG Student, Department Of ECE, LITAM Satenapalli, Guntur, Andhra Pradesh, India Chandra

More information

Segmentation and Tracking of Partial Planar Templates

Segmentation and Tracking of Partial Planar Templates Segmentation and Tracking of Partial Planar Templates Abdelsalam Masoud William Hoff Colorado School of Mines Colorado School of Mines Golden, CO 800 Golden, CO 800 amasoud@mines.edu whoff@mines.edu Abstract

More information

Adaptive Zoom Distance Measuring System of Camera Based on the Ranging of Binocular Vision

Adaptive Zoom Distance Measuring System of Camera Based on the Ranging of Binocular Vision Adaptive Zoom Distance Measuring System of Camera Based on the Ranging of Binocular Vision Zhiyan Zhang 1, Wei Qian 1, Lei Pan 1 & Yanjun Li 1 1 University of Shanghai for Science and Technology, China

More information

A Survey of Light Source Detection Methods

A Survey of Light Source Detection Methods A Survey of Light Source Detection Methods Nathan Funk University of Alberta Mini-Project for CMPUT 603 November 30, 2003 Abstract This paper provides an overview of the most prominent techniques for light

More information

Face Tracking in Video

Face Tracking in Video Face Tracking in Video Hamidreza Khazaei and Pegah Tootoonchi Afshar Stanford University 350 Serra Mall Stanford, CA 94305, USA I. INTRODUCTION Object tracking is a hot area of research, and has many practical

More information

The Kinect Sensor. Luís Carriço FCUL 2014/15

The Kinect Sensor. Luís Carriço FCUL 2014/15 Advanced Interaction Techniques The Kinect Sensor Luís Carriço FCUL 2014/15 Sources: MS Kinect for Xbox 360 John C. Tang. Using Kinect to explore NUI, Ms Research, From Stanford CS247 Shotton et al. Real-Time

More information

The Application of Image Processing to Solve Occlusion Issue in Object Tracking

The Application of Image Processing to Solve Occlusion Issue in Object Tracking The Application of Image Processing to Solve Occlusion Issue in Object Tracking Yun Zhe Cheong 1 and Wei Jen Chew 1* 1 School of Engineering, Taylor s University, 47500 Subang Jaya, Selangor, Malaysia.

More information

A Background Modeling Approach Based on Visual Background Extractor Taotao Liu1, a, Lin Qi2, b and Guichi Liu2, c

A Background Modeling Approach Based on Visual Background Extractor Taotao Liu1, a, Lin Qi2, b and Guichi Liu2, c 4th International Conference on Mechatronics, Materials, Chemistry and Computer Engineering (ICMMCCE 2015) A Background Modeling Approach Based on Visual Background Extractor Taotao Liu1, a, Lin Qi2, b

More information

DeepIM: Deep Iterative Matching for 6D Pose Estimation - Supplementary Material

DeepIM: Deep Iterative Matching for 6D Pose Estimation - Supplementary Material DeepIM: Deep Iterative Matching for 6D Pose Estimation - Supplementary Material Yi Li 1, Gu Wang 1, Xiangyang Ji 1, Yu Xiang 2, and Dieter Fox 2 1 Tsinghua University, BNRist 2 University of Washington

More information

K-coverage prediction optimization for non-uniform motion objects in wireless video sensor networks

K-coverage prediction optimization for non-uniform motion objects in wireless video sensor networks International Conference on Advanced Electronic Science and Technology (AEST 2016) K-coverage prediction optimization for non-uniform motion objects in wireless video sensor networks a Yibo Jiang, Shanghao

More information

TEGRA K1 AND THE AUTOMOTIVE INDUSTRY. Gernot Ziegler, Timo Stich

TEGRA K1 AND THE AUTOMOTIVE INDUSTRY. Gernot Ziegler, Timo Stich TEGRA K1 AND THE AUTOMOTIVE INDUSTRY Gernot Ziegler, Timo Stich Previously: Tegra in Automotive Infotainment / Navigation Digital Instrument Cluster Passenger Entertainment TEGRA K1 with Kepler GPU GPU:

More information

SURVEY PAPER ON REAL TIME MOTION DETECTION TECHNIQUES

SURVEY PAPER ON REAL TIME MOTION DETECTION TECHNIQUES SURVEY PAPER ON REAL TIME MOTION DETECTION TECHNIQUES 1 R. AROKIA PRIYA, 2 POONAM GUJRATHI Assistant Professor, Department of Electronics and Telecommunication, D.Y.Patil College of Engineering, Akrudi,

More information

Pixel-Pair Features Selection for Vehicle Tracking

Pixel-Pair Features Selection for Vehicle Tracking 2013 Second IAPR Asian Conference on Pattern Recognition Pixel-Pair Features Selection for Vehicle Tracking Zhibin Zhang, Xuezhen Li, Takio Kurita Graduate School of Engineering Hiroshima University Higashihiroshima,

More information

A Laplacian Based Novel Approach to Efficient Text Localization in Grayscale Images

A Laplacian Based Novel Approach to Efficient Text Localization in Grayscale Images A Laplacian Based Novel Approach to Efficient Text Localization in Grayscale Images Karthik Ram K.V & Mahantesh K Department of Electronics and Communication Engineering, SJB Institute of Technology, Bangalore,

More information

Open Access Surveillance Video Synopsis Based on Moving Object Matting Using Noninteractive

Open Access Surveillance Video Synopsis Based on Moving Object Matting Using Noninteractive Send Orders for Reprints to reprints@benthamscience.net The Open Automation and Control Systems Journal, 2013, 5, 113-118 113 Open Access Surveillance Video Synopsis Based on Moving Object Matting Using

More information

Human Body Recognition and Tracking: How the Kinect Works. Kinect RGB-D Camera. What the Kinect Does. How Kinect Works: Overview

Human Body Recognition and Tracking: How the Kinect Works. Kinect RGB-D Camera. What the Kinect Does. How Kinect Works: Overview Human Body Recognition and Tracking: How the Kinect Works Kinect RGB-D Camera Microsoft Kinect (Nov. 2010) Color video camera + laser-projected IR dot pattern + IR camera $120 (April 2012) Kinect 1.5 due

More information

Stereo Vision II: Dense Stereo Matching

Stereo Vision II: Dense Stereo Matching Stereo Vision II: Dense Stereo Matching Nassir Navab Slides prepared by Christian Unger Outline. Hardware. Challenges. Taxonomy of Stereo Matching. Analysis of Different Problems. Practical Considerations.

More information

INTEGRATING COMPUTER VISION SENSOR INNOVATIONS INTO MOBILE DEVICES. Eli Savransky Principal Architect - CTO Office Mobile BU NVIDIA corp.

INTEGRATING COMPUTER VISION SENSOR INNOVATIONS INTO MOBILE DEVICES. Eli Savransky Principal Architect - CTO Office Mobile BU NVIDIA corp. INTEGRATING COMPUTER VISION SENSOR INNOVATIONS INTO MOBILE DEVICES Eli Savransky Principal Architect - CTO Office Mobile BU NVIDIA corp. Computer Vision in Mobile Tegra K1 It s time! AGENDA Use cases categories

More information

Fast trajectory matching using small binary images

Fast trajectory matching using small binary images Title Fast trajectory matching using small binary images Author(s) Zhuo, W; Schnieders, D; Wong, KKY Citation The 3rd International Conference on Multimedia Technology (ICMT 2013), Guangzhou, China, 29

More information

Improvement of SURF Feature Image Registration Algorithm Based on Cluster Analysis

Improvement of SURF Feature Image Registration Algorithm Based on Cluster Analysis Sensors & Transducers 2014 by IFSA Publishing, S. L. http://www.sensorsportal.com Improvement of SURF Feature Image Registration Algorithm Based on Cluster Analysis 1 Xulin LONG, 1,* Qiang CHEN, 2 Xiaoya

More information

An Edge-Based Approach to Motion Detection*

An Edge-Based Approach to Motion Detection* An Edge-Based Approach to Motion Detection* Angel D. Sappa and Fadi Dornaika Computer Vison Center Edifici O Campus UAB 08193 Barcelona, Spain {sappa, dornaika}@cvc.uab.es Abstract. This paper presents

More information

A physically motivated pixel-based model for background subtraction in 3D images

A physically motivated pixel-based model for background subtraction in 3D images A physically motivated pixel-based model for background subtraction in 3D images M. Braham, A. Lejeune and M. Van Droogenbroeck INTELSIG, Montefiore Institute, University of Liège, Belgium IC3D - December

More information

Fast Face Detection Assisted with Skin Color Detection

Fast Face Detection Assisted with Skin Color Detection IOSR Journal of Computer Engineering (IOSR-JCE) e-issn: 2278-0661,p-ISSN: 2278-8727, Volume 18, Issue 4, Ver. II (Jul.-Aug. 2016), PP 70-76 www.iosrjournals.org Fast Face Detection Assisted with Skin Color

More information

Measurement of Pedestrian Groups Using Subtraction Stereo

Measurement of Pedestrian Groups Using Subtraction Stereo Measurement of Pedestrian Groups Using Subtraction Stereo Kenji Terabayashi, Yuki Hashimoto, and Kazunori Umeda Chuo University / CREST, JST, 1-13-27 Kasuga, Bunkyo-ku, Tokyo 112-8551, Japan terabayashi@mech.chuo-u.ac.jp

More information

Effects Of Shadow On Canny Edge Detection through a camera

Effects Of Shadow On Canny Edge Detection through a camera 1523 Effects Of Shadow On Canny Edge Detection through a camera Srajit Mehrotra Shadow causes errors in computer vision as it is difficult to detect objects that are under the influence of shadows. Shadow

More information

Research on QR Code Image Pre-processing Algorithm under Complex Background

Research on QR Code Image Pre-processing Algorithm under Complex Background Scientific Journal of Information Engineering May 207, Volume 7, Issue, PP.-7 Research on QR Code Image Pre-processing Algorithm under Complex Background Lei Liu, Lin-li Zhou, Huifang Bao. Institute of

More information

Haresh D. Chande #, Zankhana H. Shah *

Haresh D. Chande #, Zankhana H. Shah * Illumination Invariant Face Recognition System Haresh D. Chande #, Zankhana H. Shah * # Computer Engineering Department, Birla Vishvakarma Mahavidyalaya, Gujarat Technological University, India * Information

More information

Definition, Detection, and Evaluation of Meeting Events in Airport Surveillance Videos

Definition, Detection, and Evaluation of Meeting Events in Airport Surveillance Videos Definition, Detection, and Evaluation of Meeting Events in Airport Surveillance Videos Sung Chun Lee, Chang Huang, and Ram Nevatia University of Southern California, Los Angeles, CA 90089, USA sungchun@usc.edu,

More information

Embedded Linux Conference San Diego 2016

Embedded Linux Conference San Diego 2016 Embedded Linux Conference San Diego 2016 Linux Power Management Optimization on the Nvidia Jetson Platform Merlin Friesen merlin@gg-research.com About You Target Audience - The presentation is introductory

More information

An indirect tire identification method based on a two-layered fuzzy scheme

An indirect tire identification method based on a two-layered fuzzy scheme Journal of Intelligent & Fuzzy Systems 29 (2015) 2795 2800 DOI:10.3233/IFS-151984 IOS Press 2795 An indirect tire identification method based on a two-layered fuzzy scheme Dailin Zhang, Dengming Zhang,

More information

Multi-Targets Tracking Based on Bipartite Graph Matching

Multi-Targets Tracking Based on Bipartite Graph Matching BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 14, Special Issue Sofia 014 Print ISSN: 1311-970; Online ISSN: 1314-4081 DOI: 10.478/cait-014-0045 Multi-Targets Tracking Based

More information

Seminar Heidelberg University

Seminar Heidelberg University Seminar Heidelberg University Mobile Human Detection Systems Pedestrian Detection by Stereo Vision on Mobile Robots Philip Mayer Matrikelnummer: 3300646 Motivation Fig.1: Pedestrians Within Bounding Box

More information

Occlusion Detection of Real Objects using Contour Based Stereo Matching

Occlusion Detection of Real Objects using Contour Based Stereo Matching Occlusion Detection of Real Objects using Contour Based Stereo Matching Kenichi Hayashi, Hirokazu Kato, Shogo Nishida Graduate School of Engineering Science, Osaka University,1-3 Machikaneyama-cho, Toyonaka,

More information

MULTI ORIENTATION PERFORMANCE OF FEATURE EXTRACTION FOR HUMAN HEAD RECOGNITION

MULTI ORIENTATION PERFORMANCE OF FEATURE EXTRACTION FOR HUMAN HEAD RECOGNITION MULTI ORIENTATION PERFORMANCE OF FEATURE EXTRACTION FOR HUMAN HEAD RECOGNITION Panca Mudjirahardjo, Rahmadwati, Nanang Sulistiyanto and R. Arief Setyawan Department of Electrical Engineering, Faculty of

More information

Hardware Acceleration of Feature Detection and Description Algorithms on Low Power Embedded Platforms

Hardware Acceleration of Feature Detection and Description Algorithms on Low Power Embedded Platforms Hardware Acceleration of Feature Detection and Description Algorithms on LowPower Embedded Platforms Onur Ulusel, Christopher Picardo, Christopher Harris, Sherief Reda, R. Iris Bahar, School of Engineering,

More information

AAM Based Facial Feature Tracking with Kinect

AAM Based Facial Feature Tracking with Kinect BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 15, No 3 Sofia 2015 Print ISSN: 1311-9702; Online ISSN: 1314-4081 DOI: 10.1515/cait-2015-0046 AAM Based Facial Feature Tracking

More information

Efficient Acquisition of Human Existence Priors from Motion Trajectories

Efficient Acquisition of Human Existence Priors from Motion Trajectories Efficient Acquisition of Human Existence Priors from Motion Trajectories Hitoshi Habe Hidehito Nakagawa Masatsugu Kidode Graduate School of Information Science, Nara Institute of Science and Technology

More information

Pedestrian Detection with Improved LBP and Hog Algorithm

Pedestrian Detection with Improved LBP and Hog Algorithm Open Access Library Journal 2018, Volume 5, e4573 ISSN Online: 2333-9721 ISSN Print: 2333-9705 Pedestrian Detection with Improved LBP and Hog Algorithm Wei Zhou, Suyun Luo Automotive Engineering College,

More information

Introduction to behavior-recognition and object tracking

Introduction to behavior-recognition and object tracking Introduction to behavior-recognition and object tracking Xuan Mo ipal Group Meeting April 22, 2011 Outline Motivation of Behavior-recognition Four general groups of behaviors Core technologies Future direction

More information

A Semi-Automatic 2D-to-3D Video Conversion with Adaptive Key-Frame Selection

A Semi-Automatic 2D-to-3D Video Conversion with Adaptive Key-Frame Selection A Semi-Automatic 2D-to-3D Video Conversion with Adaptive Key-Frame Selection Kuanyu Ju and Hongkai Xiong Department of Electronic Engineering, Shanghai Jiao Tong University, Shanghai, China ABSTRACT To

More information