可变算子图像边缘复杂目标特征识别方法仿真  

Simulation of Complex Target Feature Recognition Method for Variable Operator Image Edge

作　　者：叶兵[1] 马伟东 YE Bing;MA Wei-dong(School of Electronic Science and Applied Physics,Hefei University of Technology,Hefei Anhui 230000,China)

机构地区：[1]合肥工业大学电子科学与应用物理学院

出　　处：《计算机仿真》2019年第10期453-457,共5页Computer Simulation

基　　金：安徽省自然科学基金资助(J2014AKZR0032)

摘　　要：图像边缘复杂目标特征识别是图像处理过程中一个重要的步骤,为了准确识别图像边缘复杂目标特征,提出基于可变算子的图像边缘复杂目标特征识别方法.将含噪图像在小波变换域中系数的聚焦特征作为研究依据,利用中值滤波方法代替高斯滤波对图像进行滤除.通过迭代算法计算图像的高、低阈值,利用数学的形式细化识别出的图像,使识别到的图像边缘较优,且能够有消除外界的干扰.与传统方法相比,所提方法具有很好的去噪效果,且能够快速、准确识别出图像边缘复杂目标特征,提升图像边缘识别效果.The recognition for complex target feature of image edge is an important step during the image processing. In order to accurately identify the complex target feature of image edge, this article puts forward a method to recognize the complex target feature of image edge based on variable operator. The focusing characteristic of coefficient of noisy image in the wavelet transform domain was taken as the research basis. The median filtering method was used to take the place of Gaussian filtering, and thus to filter the image. The iterative algorithm was used to calculate the high threshold value and low threshold value of image. The recognized image was refined by mathematical means, so that the edge of the recognized image was better and the interference of the outside could be eliminated. Compared with the traditional method, the proposed method has a good effect of noise reduction. Meanwhile, the proposed method can quickly and accurately identify complex target features at the edge of image and improve the effect of image edge recognition.

关 键 词：可变算子 图像边缘 复杂目标特征 识别方法 

分 类 号：TP315[自动化与计算机技术—计算机软件与理论]