为了能够对视频图像进行实时清晰度评价,基于Xilinx A-7型号FPGA芯片设计且实现了一种新的图像清晰度算法。首先,采用状态机设计OTSU算法计算图像的自适应阈值以分离边缘,然后应用并行处理和流水线实现改进的4方向Tenengrad梯度函数计算清晰度。实验结果表明,该设计可以在约66 ms内对640×480@30 fps格式视频图像清晰度进行评价,具有一定的灵敏度。

In order to evaluate the real-time clarity of video images,a new image clarity algorithm is designed and implemented on Xilinx A-7 FPGA chip.Firstly,a state machine design OTSU algorithm is used to calculate the adaptive threshold of the image to separate edges.Secondly,parallel processing and pipeline implementation are applied to improve the computational clarity of the 4-direction Tenengrad gradient function.The experiment results indicate that the design can complete 640×480@30 fps evaluation of the clarity of video stream images in 66 ms,with a certain degree of sensitivity.