FPGA implementation of infrared image preprocessing with cyclic lateral inhibition network

The cyclic lateral inhibition network has the functions of enhancing the contrast of image edge and highlighting the image frame in infrared imaging preprocessing. Since the processing computation complexity of the cyclic lateral inhibition network, it is important to realize the real-time preprocessing of infrared video images with cyclic lateral inhibition network. The parallel processing structure suitable for the implementation of the lateral inhibition network by hardware is proposed based on the simplified processing algorithm of the lateral inhibition network and the characteristics of hardware calculation. It is realized in FPGA by the pipeline design. The results show that the algorithm has the real-time characteristic and is synchronized with the video stream in the FPGA, and the effect of enhancing the contrast of image edge and highlighting the image frame is achieved. When it is used in real time target tracking system, the target tracking accuracy and stability are greatly increased.