Detection Of Moving Targets With Low Illumination

Moving object detection is the main research direction in the field of com puter vision.At present,most moving object detection systems process video im ages through software and under sufficient illumination,the processing cost of the system is high,the speed is slow,and it is difficult to solve the problem of target detection in low illumination environment.The video image in low ill umination environment has more noise,low contrast and brightness,which grea tly affects the detection efficiency of the system for moving objects.Aiming at the above problems,this paper designs a moving object detecti on system based on Zynq(All Programmable So C).Firstly,the hardware platfo rm and software platform of the system are built to realize the video image da ta acquisition from the camera to the output of LCD.Then the image processi ng algorithm is implemented in hardware and packaged into IP core to comple te the BD design of the system.Finally,the camera,SCCB and LCD display of the system are initialized on the software platform to complete the whole s ystem design.In order to meet the high real-time requirements of the detection system,t his paper abandons the traditional SDRAM storage module and uses VDMA co mbined with off-chip memory DDR3 to realize image frame caching.VDMA c an transmit data with high bandwidth with DDR3,which is beneficial to reduc e the delay of the system.In the design of image processing algorithm,the sp atial nonlinear median filter is used to denoise and smooth the low illuminatio n image,In order to improve the quality of video images,the gray histogram equalization algorithm is designed by Vivado HLS software to improve the bri ghtness and contrast of video images.The moving object detection adopts the i nter-frame difference method,and the mathematical morphology processing algo rithm is designed to reduce the secondary noise of the video image to further reduce the impact of the low illumination environment on the system moving object detection.The system design in this paper follows the design principles of accuracy,real-time and low power consumption.The relevant algorithms used have been simulated and verified.Through resource utilization ratio comparison and video image result analysis,the effectiveness of the algorithm in hardware implement ation is guaranteed.The test results show that the system can accurately detect moving objects in low illumination environment,and the detection effect is good,and the res ource utilization is low,which meets the design requirements of the moving ob ject detection system.