01259nam a2200133Ia 4500008004100000041001200041100001800053245004400071260004100115510084300156942000700999999001901006952010001025231221s9999    xx            000 0 und d  aEnglish  aFatima Zohra. 0aSOC For Object TrackingbRajeshwari, B.  c2018-2019bPES UniversityaBengaluru  aObject tracking has been one of the most researched topics in computer vision. A drastic improvement has been seen in the field of object tracking over past few decades. However, it still remains a complex problem to be solved. In this project a SoC is designed by integrating designed and customized IPs from library. Application program is written for the SoC to detect and track moving objects. Optical flow approach is used to detect the objects. Lucas- Kanade dense optical flow is used to obtain the information of movement of the object. Median filtering removes the unwanted noise, thresholding and morphological operations are used to track only the object of interest. Blob analysis sets the object type restrictions. The hardware implementation has been achieved on ZedBoard platform which is an All- Programmable SoC by Xilinx  cCD  c237078d237078  r2024-01-26 00:00:00l0pCD7205w2024-01-26yCD10407000aMLbMLd0000-00-00EEC/PGGRSF1V0