Research And Design Of The Image Edge Detection System Based On FPGA

The rapid and effective edge detection technology in digital image processing is the hotspot in the field of image processing.Although it can be realized by the general software processing method,it can not meet some industrial fields for edge detection because of the large amount of data and the limitation of software data processing ability Real-time requirements.FPGA has the characteristics of parallel computing and pipeline,very suitable for large amount of data and large amount of calculation but does not involve a lot of data storage processing,suitable for fast edge processing system.This paper mainly studies the threshold adaptive Sobel edge detection method based on FPGA,which has the non-maximal suppression of the edge,which can better remove the pseudo-edge to improve the precision and refine the edge width,and at the Sobel edge gradient In the threshold setting,the gradient corresponding to the maximum second derivative of the image gradient histogram curve can be used to adapt the calculated threshold to the environment change,and it has higher reliability and adaptability than the manual setting of the empirical threshold.The system of image acquisition module selection OV7725 module,according to VGA control display module and bus timing diagram,in the FPGA to complete the image acquisition,filtering and edge detection and image digital conversion all the process.System processing process,OV7725 camera through the FPGA camera control module configuration,collected YCbCr422 format image data input to the FPGA chip,after the FPGA data receiving module to convert the white image into a gray image,and then through the image processing,and finally the The image is displayed via the VGA control module.In this paper,the realization of image processing is introduced in detail,and the effect of image processing is analyzed at the end of this paper.The system design based on FPGA edge extraction processing platform is a complete hardware properties,with a high operating rate,the processing speed and the camera input video data stream the same rate,only 80.175us data stream delay,with a stable and efficient edge Extraction function,can quickly obtain the target object characteristics,can be adapted to a variety of industrial applications.