GMAT9300 Aerial and Satellite Imaging Systems

Transcription

1 GMAT9300 Aerial and Satellite Imaging Systems Semester 2, COURSE DETAILS Units of Credit 6 Contact hours 5 Class Tuesday to BUS232 Workshop Wednesday to MAT308 and CivEng Lab 201 Course Coordinator and Lecturer Lecturer Emeritus Professor John Trinder j.trinder@unsw.edu.au: CVEN 612: Dr Yincai Zhou y.zhou@unsw.edu.au INFORMATION ABOUT THE COURSE This course is a fourth year undergraduate elective 6UOC course in the B.E. (Surveying) and BE (GIS) programs and a postgraduate programs elective. HANDBOOK DESCRIPTION OBJECTIVES To present the principles of processing aerial and satellite frame and push-broom based digital images as well as close range frame images for determining accurate metric details of extracted objects for input into digital mapping and GIS databases and in close range applications. To present the principles of acquisition and processing of airborne lidar data for the determination of DEMs and information extraction TEACHING STRATEGIES A variety of teaching strategies will be included to achieve the optimal teaching and learning outcomes. Major teaching activities in this course are: 1) Weekly lectures and workshops 2) Workshops/laboratory sessions to review the content of lectures and tasks on the processing of digital photogrammetry and airborne lidar 3) Computation of block adjustment of multiple aerial images and production of orthophoto 4) Assignment on intersection computation from image coordinates on two digital images 5) Close-range photogrammetry image acquisition and processing of images 6) Processing of UAS images 7) Assignment on processing of airborne lidar data 8) 4 quizzes throughout the semester 9) Final examination The lectures will provide the foundation for the course. The workshops and laboratory classes will reinforce GMAT9300 Semester 2, Course Profile Page 1

2 the basic principles and provide relevance to the lecture materials. Private Study Review lecture slides and notes Do set problems and assignments Reflect on class problems and assignments Download materials from Moodle Consult reference books Lectures Find out what you must learn See methods that are not in the notes Follow worked examples Hear announcements on course changes Workshops Be guided by demonstrators Practice solving set problems Ask questions Assessments (4 quizzes throughout Demonstrate your knowledge and skills learnt the semester, final examination, assignments and reports) Demonstrate higher understanding and problem solving Laboratory Work Hands-on work, to set studies in context The basic tools will be: o ERDAS Imagine software for digital photogrammetric workstation. o Australis Close Range Photogrammetry software o Pix4D software for UAS image processing o Airborne lidar data processing using QT software. EXPECTED LEARNING OUTCOMES By the end of this session you should be able to: The student will gain an appreciation of the advantages and disadvantages of the imaging systems and airborne lidar for extraction of information for digital mapping and GIS databases. The student should learn how the imaging and airborne lidar systems can provide accurate metric descriptions of terrain details, as well as the manual and automatic extraction processes used for aerial, satellite and close range imaging systems. This course is designed to address the learning outcomes and the corresponding Engineers Australia Stage 1 Competency Standards for Professional Engineers as shown below. Learning Outcome Explain the principles of determining geometry of terrain features and close range objects from digital data based on frame and push-broom imaging Understand and experience processing of analytical photogrammetry for multiple overlapping frame and pushbroom aerial and UAS images by aerial photogrammetric bundle adjustment Understand the acquisition and processing of airborne lidar data and its potentials for extraction of DEMs and other terrain information Understand the calibration, processes and products of mapping from aerial, unmanned aerial systems (UAS) and satellite images EA Stage 1 Competencies PE1.1, PE1.2, PE1.3 PE1.1, PE1.2, PE1.3 PE1.1, PE1.2, PE1.3 PE2.1, PE2.2, PE3.4 For each hour of contact it is expected that you will put in at least 1.5 hours of private study. GMAT9300 Semester 2, Course Profile Page 2

3 ASSESSMENT SUMMARY Overall rationale for assessment components and their association with course objectives. The final grade for this course will be based on the sum of the scores from each of the assessment tasks. Four (4) multiple choice quizzes as listed in the schedule 20% Successful completion of practical close range exercise 5% Successful completion of analytical photogrammetry using 5% ERDAS software Assignment on computation of intersection from 10% image coordinates on 2 photos Successful completion of and UAS computation exercise 5% Successful completion of airborne lidar exercise 5% Final Exam 50% Total 100% Note: The lecturer reserves the right to adjust the final scores by scaling if agreed by the Head of School. Details of each assessment component, the marks assigned to it, the criteria by which marks will be assigned, and the dates of submission are set out below. GMAT9300 Semester 2, Course Profile Page 3

4 ASSESSMENT OVERVIEW Item Length Weighting Learning outcomes assessed 2.Quizzes Quiz 1 20 min 5% Data acquisition systems, image geometry, collinearity eqns Quiz 2 20 min 5% Block adjustment of frame and pushbroom images, DEM & orthophoto computations Quiz 3 20 min 5% Processing of airborne lidar data Quiz 4 20 min 5% Close range photogrammetry and acquisition and processing of UAS images Assessment Criteria Due date and submission requirements Correct answers to quiz questions Correct answers to quiz questions Correct answers to quiz questions Correct answers to quiz questions Deadline for absolute fail Marks returned 8 August 5 22 August 5 12 September 5 4 October 5 3. Assignments Assignment 2-4 hours 5% Ability to process Submission of output from block 18 August 1 September 5 1a block adjustment of adjustment by ERDAS software aerial images revealing RMSE derived from residuals of image coordinates < 50 µm and RMSE residuals at ground control < 0.2 m. The submission should be accompanied by a brief report demonstrating an understanding of the quality of block adjustment. Assignment 2-4 hours 5% Completion of Submission of output from DEM 1 September 15 5 GMAT9300 Semester 2, Course Profile Page 4

5 1b computation of DEM and orthophoto Assignment 2 20 hours 10% Intersection computation Assignment 3 3 hours 5% Close range imaging and computation Assignment 4 3 hours 5% Processing of UAS image block Assignment 5 3 hours 5% Processing of airborne lidar data using QT software computation and orthophoto by ERDAS software. The submission should be accompanied by a brief report on visual tests of the quality of the DEM computation and the orthophoto. Submission should be a report which must include results of the computations, demonstrating each iteration and the final coordinates of the 3 object points, together with the residuals in the image coordinates in µm Photography of selected site and successful processing of images using Australis software. Submission to include a table of output from the exterior orientation computations, the residuals for the measured distances used for scaling of the object and a 3D display of the object Submission must include details of the successful processing of UAS images using Pix4D software. Submission: Completed tasks using QT software as specified in the instructions, including the production of a video. 4. Final Exam 50% Complete course Correct overall answers to examination questions September 8 September 22 September 22 September 6 October 6 October 20 October 20 October 3 November During period Exams GMAT9300 Semester 2, Course Profile Page 5

6 DESCRIPTION OF ASSIGNMENTS ALL SUBMISSIONS SHOULD BE MADE THROUGH MOODLE Exercise 1 (in 2 parts): Completion of an orthophoto by ERDAS software, issued on: 7 August due for completion on: 18 August. Depending on the number of students in the class, the 2 components of this exercise can be undertaken individually or by groups of 2 students. A comprehensive exercise guide will be provided so that students can undertake the exercise under minimum supervision. Some aspects may need to be undertaken in batch mode because of the time involved in computing the DEM and orthophotos. The exercises can also be undertaken outside the prescribed lab times. Commencing and completion times for this exercise are indicative only, but it is advisable to adhere to these times so that Exercise 2 can be completed on time. Exercise 1 Part 1: The exterior orientation of a block of digital aerial images will be processed individually or in groups of 2 students who undertook Exercise 1. Students will also be required to observe and extract image coordinates of 3 points, together with 3D object coordinates and the exterior orientations X c, Y c Z c ω, φ, κ of the 2 images on which the points appear. Information extracted from Exercise 1 Part 1 is required for Exercise 2. Submission: Output from block adjustment by ERDAS software revealing RMSE derived from residuals of image coordinates < 50 µm and RMSE residuals at ground control < 0.2 m. The submission should be accompanied by a brief report demonstrating an understanding of the quality of block adjustment. Exercise 1 Part 2: The second component of this exercise will involve the computation of a DEM over an area of the images and the production of an orthophoto. Submission: Final version of the orthophoto and a brief report with comments on visual tests of the quality of the DEM and the orthophoto. Exercise 2: Computation of object coordinates by intersection: can be commenced after completion of Exercise 1 part 1 from 16 August due on: 8 September. The exercise involves the computation of the 3D object coordinates of the 3 points that were observed in Exercise 1 part1. The purpose of this exercise is for students to appreciate the iterative computations that are necessary in solving a non-linear set of 4 equations with 3 unknowns. Submission: A report which must include results of the computations demonstrating each iteration, and the final coordinates of the 3 object points, together with the residuals in the image coordinates in µm Exercise 3: Close range survey from photos taken on campus: issued on: 12 September, due for completion on: 22 September. This exercise provides students with experience in taking digital photos for close range photogrammetric measurement of objects and processing the images on Australis software, a state-of-the art packaged developed in Australia. The results of the exercise will be judged by the quality of the definition of the photographed objects as well as the output displaying the quality of the orientation of the images. Submission: Must include a table of output from the exterior orientation computations, the residuals for the measured distances used for scaling of the object and a 3D display of the object as well as a brief description of the processing steps and comments on your close range image acquisition and image processing. What can be improved if you carry out this project once again? Exercise 4: Computation of block adjustment of UAS images using Pix4D software issued on 19 September, due for completion on 6 October. Submission: must include details of the successful processing of UAS images using Pix4D software including the final report generated by Pix4D, screenshot of map view with flight paths and image locations, 3D view of ray cloud with camera locations, and other supporting screenshots. A brief description of the processing steps and comments on GCP and point cloud accuracies on your UAS image processing. Exercise 5: Processing of airborne lidar data using QT software: Issued on 11 October, due 20 October. Submission: Completed the tasks and answer the questions in the instructions for the use of QT software, including the production of a video. GMAT9300 Semester 2, Course Profile - 6

7 COURSE SCHEDULE The table of lectures and workshops for each week. Lectures will be given by Prof Trinder unless otherwise stated. SEMESTER 2, Week Beginning Topic Workshops Date 1 24/7/ 2 31/7/17 3 7/8/ /8/ /8/ /08/17 Introduction to subject. Definitions of photogrammetry and remote sensing, applications. Short history. Chapter 1: Sensors and Platforms for Acquisition of Aerial and Satellite Image Data Design of digital aerial cameras, overlaps of aerial photography. Characteristics of aerial and satellite images. Chapter 2: Fundamental Geometry of Single and Overlapping Images. Geometry of whiskbroom imaging systems. Direct and indirect stereovision, instrumentation, epipolar geometry. Principles of the floating mark, elevation determination by parallax equations Photograph coordinate systems Chapter 3: Analytical photogrammetry & Block adjustment Collinearity equations, Principles of analytical photogrammetry; Collinearity equations for computation of block adjustment 1 st quiz Chapter 3: Block adjustment of multiple aerial frame photographs; orientation of pushbroom aerial scanner images. Chapter 3: Orientation of pushbroom scanner satellite images. Examples of questions on equations for analytical photogrammetry. Workshop (1 hour): Computation of intersection. 2 nd quiz Chapter 4 Digital Image Processing for Elevation Determination and Orthophotos - image matching: geometric correction of digital images for orthophoto production; resampling; digital photogrammetric software Chapter 5. Processing of Airborne Laser Scanning (ALS) Lidar Data Characteristics of lidar data, filtering, errors, extraction of DEMs and other features Lecture (MAT308): Chapter 1 (cont) Whiskbroom sensors, radar image sensors, airborne lidar. Chapter 2: Fundamental Geometry of Single and Overlapping Images. Central projection and orthographic projection. Photograph scale, tilt and relief displacement. Workshop: Prof Trinder (MAT308): Applications of collinearity eqns. Worked examples on image coordinates and exterior orientation Commence Exercise 1 part 1: Lab exercise (CE201): Exterior orientation of multiple aerial photos and extraction of image and object coordinate of features Exercise 1 part 1 (cont): Lab exercise (CE201): Exterior orientation of multiple photos and extraction of image and object coordinates of points for intersection computation. Commence Exercise 2 Commence Exercise 1 part 2: Lab exercise (CE201): Production of DEM and Orthophoto Exercise 1 part 2 (cont): Lab exercise (CE201): Production of DEM and Orthophoto GMAT9300 Semester 2, Course Profile - 7

8 7 4/9/17 Lecture: Chapter 5 Processing of Airborne Laser Scanning (ALS) Lidar Data Characteristics of lidar data Prof Trinder: (MAT308): Workshop: Assessment of accuracy of information extraction from aerial images Submit Exercise 2 (Friday 8 Sept) 8 11/9/17 3 rd quiz Lecture by Dr Zhou: Close range photogrammetry data acquisition and processing. Workshop: Instructions on Acquisition of closerange photogrammetry images and software processing Commence Exercise 3 Dr Zhou: Lab exercise (CE201): Processing of close range photos with Australis 9 18/9/17 Lecture by Dr Zhou: Acquisition of UAS images and processing Commence Exercise 4: Lab exercise (CE201) Computation of UAS images using Pix4D software Submit Exercise 3 (Friday 22 Sept) Break 10 2/10/17 4 th quiz Chapter 6: Mapping procedures from aerial and satellite images Yincai absent Submit exercise 4 (Friday 6 Oct) Commence Exercise 5: Lab exercise (CE201): Operations of airborne lidar software 11 9/10/17 No lecture Yincai absent Lab exercise (CE201): Operations of airborne lidar software 12 16/10/ /10/17 Chapter 6: Mapping procedures from aerial and satellite images (cont) Submit Exercise 5 (Friday 20 Oct) Workshop (MAT308): Semiautomatic and automatic semantic information extraction from images Workshop (MAT308): Revision of course RELEVANT RESOURCES Notes and slides will be provided for all chapters except the final topic on information extraction. These notes should be reviewed before each lecture. The following texts have been listed by the Bookshop, but would have to be ordered for purchase. GMAT9300 Semester 2, Course Profile - 8

9 o Photogrammetry Geometry from Images & Laser Scans Karl Kraus Karl (translated from German by I. A. Harley and Stephen Kyle) de Gruyter Textbook 2007 (a good text but uses different symbology than in the lectures) o Geoinformation 2 nd edition by G. Konecny CRC Press 2014 (This is the most up-to-date text covering photogrammetry and an excellent text. It also covers Remote Sensing and GIS) For further reading and availability (in UNSW Library). Manual of Photogrammetry, Ed C. McGlone, published by American Society for photogrammetry and Remote Sensing 6 th edition (latest) but expensive. DATES TO NOTE Refer to MyUNSW for Important Dates available at: PLAGIARISM Beware! An assignment that includes plagiarised material will receive a 0% Fail, and students who plagiarise may fail the course. Students who plagiarise are also liable to disciplinary action, including exclusion from enrolment. Plagiarism is the use of another person s work or ideas as if they were your own. When it is necessary or desirable to use other people s material you should adequately acknowledge whose words or ideas they are and where you found them (giving the complete reference details, including page number(s)). The Learning Centre provides further information on what constitutes Plagiarism at: ACADEMIC ADVICE For information about: Notes on assessments and plagiarism, School policy on Supplementary exams, Special Considerations, Solutions to Problems, Year Managers and Grievance Officer of Teaching and Learning Committee, CEVSOC. Refer to Academic Advice on the School website available at: GMAT9300 Semester 2, Course Profile - 9