Projects:2020s2-7292 StableEyes
Introduction The stable eye project tends to develop a video process system to track and stabilize targets in unstable footage by mix combine the rapid manual tracking and Auto-detection tracking. This system will be implemented to a windows 10 executable application. This application will generate track vector file and processed footage.
Project Team Juntao Miao a1731105 Qing Yang a1715206
Supervisors Matthew Sorell Richard Matthews
Aim and Motivation Develop a video process system to track and stabilize targets in unstable footage This system will be implemented to a desktop executable application This application will be generating track vector file and processed footage.
Background In modern society, it is become more common to use smart phones to take criminal evidence. Therefore, law enforcement agencies need to improve the ability to use smart devices’ video to crack down on criminals. However, the quality of many video evidence received by law enforcement agencies is poor, which will affect the tracking of objects or people in the video and affect the performance of video stabilizer.
Structural design
Evaluation we find a video and manual mark a people. Then we let our code to learn its features. Then we go through the next frame to evaluate the detect performance. There are two involved auto-detection algorithm which are surf and Meanshift respectively.
Result We put our manual mark result as the refer standard. By compared it similarity we found the surf detection is almost 90 percent similar with our standard however the Mean shift only like 25 percentage which are poor performance than surf in this scenario. And this evaluate method is absolutely help in our project, this test case is just an example.
References [1] Video Evidence A Law Enforcement Guide to Resources and Best Practices. U.S Department of Justice, 2020, p. 1. [2] G. Ciaparrone, F. Luque Sánchez, S. Tabik, L. Troiano, R. Tagliaferri and F. Herrera, "Deep learning in video multi-object tracking: A survey", Neurocomputing, vol. 381, pp. 61-88, 2020. Available: 10.1016/j.neucom.2019.11.023. [3]A. Ajith, L. Jaime and F. John, Advances in Computing and Communications. kochi, 2020, pp. 349-357. [4] G. Cao, L. Huang, H. Tian, X. Huang, Y. Wang and R. Zhi, "Contrast enhancement of brightness-distorted images by improved adaptive gamma correction", Computers & Electrical Engineering, vol. 66, pp. 569-582, 2018. Available: 10.1016/j.compeleceng.2017.09.012.