Design And Implementation Of Aricultural Field Video Image Real-time Segmentation Based On FPGA

With the continuous expansion of the scale of image data,image processing speed and real-time performance of digital image segmentation algorithms are increasingly demanding.For some algorithms requiring milliseconds real-time processing,it is difficult for traditional software platforms to complete the corresponding operations within the specified time.The Field Programmable Gate Array(FPGA)has a rich register resource and pipeline structure,which can process a large amount of image data in real time,greatly improving the speed of image processing.This thesis designs and implements a video image acquisition processing and display system for an agricultural robot.The system can collect real-time field video images in complex environments,and improve image super-green area extraction,Canny edge detection,OTSU threshold segmentation and median filtering algorithm for FPGA hardware environment,so as to quickly and accurately capture the captured image data.Split,its split image can be displayed on the display through the interface.In this thesis,the image greenness extraction,Canny edge detection,OTSU threshold segmentation and median filtering algorithm are optimized by parallel pipeline.The four modules in the image threshold segmentation algorithm are designed in parallel,which effectively improves the image processing speed.The experimental results show that the segmentation algorithm used in this thesis only needs 2.44 ms for image segmentation with resolution of 640×480,and only 16.01 ms for image segmentation with resolution of 1920×1080.The implementation speed is about PC tested in this thesis.Machine software achieves a dozen times faster.The occupation trigger FF,the use of the lookup table LUT and the block register BRAM are 2824,4625 and 4 KB,respectively,and the algorithm design is implemented with fewer resources.The system experiments show that the FPGA-based field video image real-time segmentation system can complete the field video image acquisition,buffering,filtering and segmentation,and display the processed results on the display,and the algorithm implemented by FPGA is more efficient than the traditional software method.High,meeting the expected requirements.