为了改善红外图像的视觉效果, 介绍了一种基于 FPGA的红外图像自适应分段线性拉伸增强算法。该算法根据红外视频图像帧间的高度相关性和图像的特点, 通过统计红外图像多级均值的方法估算图像灰度阈值, 自适应地将图像分成背景灰度区间、过度灰度区间和目标灰度区间, 最后对各部分区间采取不同的线性变换, 获得最佳的视觉效果。该算法已在自主研发的红外手持观测仪上得到了应用, 占用硬件逻辑资源少, 实时性强, 具有一定的应用价值。

To improve the visual effect of infrared image, an adaptive piecewise linear enhance algorithm based on FPGA is introduced. According to high correlation between the frames of infrared video and the feature of infrared image, this algorithm calculates the threshold from statistics of multistage mean, adaptively divides image into background grayscale section, interim grayscale section and target grayscale section, then applies different linear grayscale transformation to every section in order to get the best visual effect. It has been proved that in infrared observation system developed ourselves, this algorithm cost less hardware resource, and fully satisfied the real-time, have a certain value of use.