摘要
Abstract
Due to the influence of laser coherence,speckle noise often exists in lidar images,which makes the image appear as if it is covered with a layer of random particles,reducing the sharpness of the image.Therefore,a lidar image enhancement algorithm considering morphological features is proposed.An incomplete wavelet tree structure is used to decompose image sub bands,and the average energy and ratio of their horizontal and vertical boundary features are calculated to obtain the morphological features of each sub band.Based on the morphological feature results,edge and directional information are used to guide morphological operations,and lidar image filtering is completed by erosion and dilation processing.Morphological feature filtering can deal with specific texture structures in images accordingly,improve overall visual effects and preserve key morphological features.The fitness function is constructed with the cuckoo search algorithm,the parameters of the Sigmoid function is adjusted,and the color of the output image is restored by combining the linear scaling factors.On the basis of the above,the adaptive enhancement of lidar images is achieved.The experimental results show that the morphological feature filtering method can achieve adaptive enhancement of lidar images,and always keep information entropy above 9.5.To sum up,the proposed method can enrich the detail information of lidar images and improve visual sharpness.关键词
形态特征滤波/激光雷达图像/纹理特征/自适应增强/布谷鸟搜索算法/小波树形结构Key words
morphological feature filtering/lidar image/texture feature/adaptive enhancement/cuckoo search algorithm/wavelet tree structure分类
信息技术与安全科学
