International Journal of Modern Trends in Engineering and Research e-issn No.: , Date: April, 2016

Size: px

Start display at page:

Download "International Journal of Modern Trends in Engineering and Research e-issn No.: , Date: April, 2016"

Giles Patterson
5 years ago
Views:

International Journal of Modern Trends in Engineering and Research www.ijmter.com e-issn No.

N. Sapkal COE, Nashik, prachigaidhani@gmail.com 2 Department of Electronics and Telecommunication, KCT s Late G.N. Sapkal COE, Nashik, bagalsaheb@rediffmail.

1 International Journal of Modern Trends in Engineering and Research e-issn No.: , Date: April, 2016 Implementation of Sketch Based and Content Based Image Retrieval Prachi A. Gaidhani 1, Prof. Bagal S. B. 2 1 Department of Electronics and Telecommunication, KCT s Late G.N. Sapkal COE, Nashik, prachigaidhani@gmail.com 2 Department of Electronics and Telecommunication, KCT s Late G.N. Sapkal COE, Nashik, bagalsaheb@rediffmail.com Abstract In this paper we propose a novel approach to combine Sketch Based and Content Based Image retrieval. The database of the images is growing rapidly and there is huge demand in the enhancement of the retrieval of the images. Color, shape, texture, spatial layout are the main attributes to represent as well index the images. These features of images are extracted to check similarity between the images. Generation of special query is the chief trouble of content based image retrieval. By using different algorithms the similarity between the extracted features from the images are checked. Sketch based image retrieval is efficient and main method which are not necessarily having a high skill to draw the query sketch. To overcome these disadvantages we proposed combination of Sketch Based and Content Based Image retrieval. Experimental results show that proposed system is much better than the single systems. Keywords- Sketch based image retrieval, Content based image retrieval, Feature extraction, Line segment based descriptor, object boundary selection. I. INTRODUCTION Sketch Based Image Retrieval (SBIR) and Content Based Image Retrieval (CBIR) problems are discuss by many researchers in different ways and numbers of techniques are proposed to retrieve images effectively from the databases. All the existing systems are independent like some retrieve the content based images and some retrieve sketch based images. To improvise the search efficiency we come up with a effective way by combining the sketch and content based techniques. An image retrieval system provides an effective way to access, browse, and retrieve a set of like images in the real-time applications. We can easily recognize objects from other people s sketches, and this form of expression is arguably the most common interaction tool for people who speak different languages. Compared with keywords, a sketch is generally more descriptive, breaking down the language barrier. Content-based image retrieval (CBIR) is the application of computer sight to the image retrieval trouble, so as to the hard of searching for digital image in big databases. "Content-based" means with the purpose of explore will examine the literal contents of the image. Here the term content in this context might submit to shade, figure, surface, or some other information that can be derived from the image itself [1]. Implementing SBIR and CBIR together will increase the searching efficiency of the system. Several approaches for SBIR have been proposed. However, to attain interactive query reply, it is impossible to compare the sketch to all images in the database right away. Instead, descriptors are extracted in a preprocess and stored in a data structure for fast access. Very commonly, the descriptors are interpreted as points in a high-dimensional space and finding close matches means searching for nearest neighbors in this space. Researchers have developed many offline as well as online CBIR systems using a combination of two or more image features and some have been successful in retrieving similar images with All rights Reserved

2 precision and recall values. Several CBIR systems such as VIR image engine from Virage Inc., IBM s QBIC and Excalibur from Excalibur technologies are available as commercial packages. Many systems such as Visual SEEK, Surf image, Photo book, Chabot, Netra and MARS have been developed on experimental basis by academic institutions and researchers to demonstrate the feasibility of new techniques. Working demonstrations of many of these are available on the web. II. LITERATURE SURVEY Many research attempts have been committed in addressing the Content Based Image Retrieval (CBIR) problem [2] ans Sketch Based Image Retrieval (SBIR). Various methods have been proposed in the literature for sketch-based and content based Retrieval. The main aspect of image retrieval system is to offer an efficient way to access, browse, and find similar images in the real-time applications. 2.1 Content Based Image Retrieval (CBIR): (CBIR) is a method that is used to look at image features like (color, shape, texture) to find a query image from database. The difficulties of CBIR lie in reducing the differences of contents based feature and the semantic based features. This problem in giving effective retrieval images and channelize the researchers to use (CBIR) system,to take global color and texture features to achieve, the good retrieval, where others used local color and texture features[3].the method in [4] presented the holistic representation of spatial envelop with a very low dimensionality for making the incident image. This approach presented an outstanding result in the scene categorization. The method in [5] proposed a modern approach for image classification with the open field design and the concept of over-completeness methodology to achieve a preferable result. As reported in [5], this method achieved the best classification performance with much lower feature spatiality compared to that of the former schemes in image classification task. Tiwari et al developed a CBIR system [PATSEEK] for US based patent database as a patent always consists of an image along with textual information. For similarity search [6] the user need to enter keywords along with the query image that might appear in the text of patents. Krishnan et al developed CBIR based on color, based on the rife colors in the foreground image which gives only the semantics of the image. Dominant color identification by using foreground objects alone is able to retrieve number of similar images considering the foreground color irrespective of size. Higher average precision and recall rates compared to the traditional Dominant Color method were obtained successfully [7].In another system the image is represented by a Fuzzy Attributed Relational Graph (FARG) that describes each object in the image, its attributes and spatial relation. The texture and color attributes are computed in a way that model the Human Vision System (HSV) [8]. 2.2 Sketch Based Image Retrieval (SBIR): Sketch-based image retrieval (SBIR) is a relevant means of querying large image databases. All of researches focus on how to solve the gap between sketch and image matching problem. The Sketch-based image retrieval (SBIR) was introduced in QBIC [20] and Visual SEEK [9] systems. In these systems the user draws color sketches and blobs on the drawing area. The images were divided into grids, and the color and texture features were determined in these grids. The applications of grids were also used in other algorithms, for example in the edge histogram descriptor (EHD) method [10]. The disadvantage of these methods is that they All rights Reserved 687

3 not invariant opposite rotation, scaling and translation. Lately the development of difficult and robust descriptors was emphasized. Another research approach is the application of fuzzy logic or neural networks. In these cases the purpose of the investment is the determination of suitable weights of image features [11].Early sketch based image retrieval systems were typically driven by queries comprising blobs of color or predefined texture [12] [13]. Later systems explored shape descriptors [14] and spectral descriptors such as wavelets [15]. Eitz et al. [16] introduced a grid based approach to shape retrieval, dividing the image into regular grids and locate photos using sketched depiction of object shape. Sciascio et al. [14] investigate extracting shape feature in photo-realistic images using image segmentation.ma et al. [17] proposed a Poisson-based HOG and organized the codebook in a hierarchical vocabulary tree. To better measure visual similarity, Shrivastava et al. [18] learned the datadriven uniqueness (weights) for blocks of HOG. Gong et al. [19] combined HOG and MLBP [20] for face image retrieval. In addition to HOG based descriptors, Eitz et al. [21] utilized shape context [22] to perform retrieval. III. PROPOSED SYSTEM Figure 1. General Architecture of sketch and content based image retrieval system Figure shows a typical architecture of a sketch and content based image retrieval system. In this system the query sketch was taken by translating original image into sketch image. Then we propose a new line segment-based descriptor named histogram of line relationship (HLR).our line segment-based descriptor HLR has two advantages. First, it focuses on directly capturing higher level information. Since the object shape is determined by these relationships, our HLR is better able to describe the sketches/object boundaries. Second, since our HLR is a line segment-based descriptor, it is able to selectively capture a subset of neighboring line segments rather than capturing all of them, which makes it quite flexible and serve as the basis for noise impact reduction. The sampling strategy has a significant influence on performance. Noisy edges widen the appearance gap between sketch images and photo-realistic images and subsequently degrade retrieval performance. However, the human visual system is not usually impacted by these noisy edges since humans can differentiate the shaping edges and noisy edges based on their inference ability. Inspired by this observation, we propose a novel object boundary selection algorithm to predict the shaping edges. To the best of our knowledge, this object boundary selection algorithm is the first method to aim to reduce the impact of noisy edges in a SBIR system. Then images are retrieved for SBIR system. These digital images are given to content based image retrieval system(cbir).content based means that search will analyze the actual content of the image. The term content in this context might refer colors, shapes, texture or other information that can be derived from the image All rights Reserved 688

4 The CBIR system uses ordered-dither block truncation coding (ODBTC) for the generation of image content descriptor. In the encoding step, ODBTC compresses an image block into corresponding quantizers and bitmap image. Two image features are proposed to index an image, namely, color co-occurrence feature (CCF) and bit pattern features (BPF), which are generated directly from the ODBTC encoded data streams without performing the decoding process. Then images are retrieved for CBIR system. CBIR systems can make use of relevance feedback; Relevance feedback retrieval systems prompt the user for feedback on retrieval results and then use this feedback on subsequent retrievals with the goal of increasing retrieval performance. IV. RESULTS 4.1 Precision and Recall graph diagram: Figure 2. Performance Graph These are approximately estimated results for precision and recall graph calculated from the formula as follows:- Precision = No. of relevant images retrieved / Total no. of images retrieved Recall = No. of relevant images retrieved / Total no. of relevant images retrieved Table 1. Comparison table Recall Existing Proposed V. CONCLUSION In the proposed system we have merged Content based and Sketch based techniques. The results shown in the result section shows the comparison of the existing system with the proposed systems. 0.5% of efficiency is increased as per the data table discussion. REFERENCES [1] SamyAit-Aoudia, RamdaneMahiou, BillelBenzaid, YACBIR: Yet another Content Based Image Retrieval system, IEEE R. naraine," Massive ddos attack hit dns root servers, All rights Reserved 689

5 [2] J.M. Guo and M.-F. Wu, Improved block truncation coding based on the void-and-cluster dithering approach," IEEE Trans. Image Process., vol. 18, no. 1, pp , Jan [3] R. Datta, D. Joshi, J. Li, and J. Z. Wang, Image retrieval, ACM Comput. Surv., Vol. 40, no. 2, Apr. 2008, pp [4] A. Oliva and A. Torralba, Modeling the shape of the scene: A holistic representation of the spatial envelope, Int. J. Comput. Vis., vol. 42, no. 3, pp , [5] Y. Jia, C. Huang, and T. Darrell, Beyond spatial pyramids: Receptive field learning for pooled image features, in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), Jun. 2012, pp [6] A. Tiwari and V. Bansal, PATSEEK: Content Based Image Retrieval System for Patent Database, Proceedings of international conference on electronic business, pp [7] N. Krishnan, M.S. Banu and C. CallinsChristiyana, Content Based Image Retrieval Using Dominant Color Identification Based on Foreground Objects, International Conference on Computational Intelligence and Multimedia Applications, Vol. 3, pp , December [8] HebaAboulmagd Ahmed, Neamat El Gayar, HodaOnsi A New Approach in Content-Based Image Retrieval Using Fuzzy Logic INFOS2008, [9] J.R. Smith, and S.F. Chang, VisualSEEK: a fully automated contentbased image query system, ACM Multimedia 96, pp , [10] M. Eitz, K. Hildebrand, T. Boubekeur, and M. Alexa, An evaluation of descriptors for large-scale image retrieval from sketched feature lines, Computers and Graphics, vol. 34, pp , October [11] D.G. Lowe, Distinctive image features from scale-invariant keypoints, International Journal of Computer Vision, vol. 60, pp , [12] J. Ashley, M. Flickner, J. L. Hafner, D. Lee, W. Niblack, and D. Petkovic, The query by image content (QBIC) system, in SIGMOD Conference, 1995, p [13] J.R. Smith and S.-F. Chang, Visualseek: a fully automated content-based image query system, in ACM Multimedia, New York, NY, USA, 1996, pp , ACM. [14] E. Sciascio, M. Mongiello, and M. Mongiello, Content-based image retrieval over the web using query by sketch and relevance feedback, in In Proc. of 4 th Intl. Conf. on Visual Information Systems, 1999, pp [15] C. E. Jacobs, A. Finkelstein, and D. H. Salesin, Fast multiresolution image querying, in Proc. ACM SIGGRAPH, Aug.1995, pp [16] M. Eitz, K. Hildebrand, T. Boubekeur, and M. Alexa, A descriptor for large scale image retrieval based on sketched feature lines, in SBIM, 2009, pp [17] C. Ma, X. Yang, C. Zhang, X. Ruan, M.-H. Yang, and O. Coporation, Sketch retrieval via dense stroke features, in Proc. Brit. Mach. Vis. Conf., 2013, vol. 2, pp [18] A. Shrivastava, T. Malisiewicz, A. Gupta, and A. A. Efros, Data-driven visual similarity for crossdomain image matching, ACM Trans. Graph., vol. 30, no. 6, pp. 154:1 154:10, Dec [19] D. Gong, Z. Li, J. Liu, and Y. Qiao, Multi-feature canonical correlation analysis for face photo-sketch image retrieval, in Proc. 21st ACM Int. Conf. Multimedia, New York, NY, USA, 2013, pp [20] T. Menp and M. Pietikinen,, J. Bigun and T. Gustavsson, Eds., Multiscale binary patterns for texture analysis, in Image Analysis, ser. Lecture Notes Comput. Sci.. Berlin, Germany: Springer-Verlag, 2003, vol. 2749, pp [21] M. Eitz, K. Hildebrand, T. Boubekeur, and M. Alexa, Sketch-based image retrieval: Benchmark and bag-of- features descriptors, IEEE Trans. Vis. Comput. Graph., vol. 17, no. 11, pp , Nov [22] S. Salve and K. Jondhale, Shape matching and object recognition using shape contexts, in Proc. 3rd IEEE Int. Conf. Comput. Sci. Inf. Technol., 2010, vol. 9, pp. 471 All rights Reserved 690

Image Retrieval Based on its Contents Using Features Extraction

Image Retrieval Based on its Contents Using Features Extraction Priyanka Shinde 1, Anushka Sinkar 2, Mugdha Toro 3, Prof.Shrinivas Halhalli 4 123Student, Computer Science, GSMCOE,Maharashtra, Pune, India