System Block Diagram
Actual Tracking Result
High Speed CMOS Sensor Target Detection and Tracking System
Target detection and tracking system was developed using simple algorithms for high speed embedded applications. Target detection was done through profiling and localized profiling of the image in the X and Y axis. Once detected, the FPGA system only transmits 20×20 pixel image chunks around the target for sub-pixel analysis. The Cypress LUPA-300 was used as the image sensor. Fixed Pattern Noise correction algorithms were implemented in Linux since the LUPA-300 does not have FPN correction on chip. To test the accuracy of the system, a stepper motor actuated jig was created to simulate movements in the X and Y axis. The motorized jig is capable of movement resolution up to 0.0254mm/step. The FPGA system was able to achieve an average processing time of 5msec per frame or around 200 frames per second. The tracking accuracy of the system is around 0.1mm to 0.2mm.