嵌入式激光图像的边缘检测与分析  

Edge Detection and Analysis of Embedded Laser Image Influence of Process Parameters on Laser Cladding Crack Defects

作　　者：高一文[1] 刘礼培[1] 周虹君[1] 

机构地区：[1]重庆文理学院,重庆402160

出　　处：《激光杂志》2016年第7期20-22,共3页Laser Journal

基　　金：国家自然科学基金资助项目(61271452)

摘　　要：边缘保留了嵌入式激光图像的重要信息,可以描述目标物体的轮廓,为了提高嵌入式激光图像边缘的检测精度,提出一种剪切波和Canny算子相融合的嵌入式激光图像边缘检测算法(Shearlet-Canny)。首先采用Canny算子对嵌入式激光图像边缘进行检测,根据图像特征确定高、低阈值,然后采用剪切波变换对嵌入式激光图像进行多个方向的分解,并对边缘检测结果进行重构和融合,最后实验结果结明,Shearlet-Canny完整保留了嵌入式激光图像的边缘信息,抗噪性能好,检测结果优于传统嵌入式激光图像边缘检测算法,具有更好的实用性。Edges retain the important information of embedded laser image, which can describe outline of target object, in order to improve detection precision of embedded laser image edge, an embedded laser image edge detection algorithm （Shearlet-Canny） based on Shearlet wave and Canny operator is proposed. First of all, Canny operator is used to detect edge of embedded laser image, and level of the threshold is automatically determined according to the characteristics of the image, and then Sheadet transform is used to decompose embedded laser image in multiple directions, and the result is fused with the edge detection result, finally, the experimental results show that Shearlet-Canny can retain the edge information of the whole embedded laser image, has good anti-noise performance, the detection result is better than traditional edge detection algorithms, and has better practicability.

关 键 词：激光图像 嵌入式技术 图像检测 仿真测试 

分 类 号：TN21[电子电信—物理电子学]