%0 Journal Article
%T 高斯差分滤波显著性的刀具磨损检测<br>Tool Wear Detection using Gauss Filter Saliency
%A 管声启
%A 洪奔奔
%A 梁洪
%A 王立中
%J 机械科学与技术
%D 2018
%X 为了提高刀具磨损区域检测准确性，本文在研究刀具磨损区域特点的基础上，提出了一种新的刀具磨损检测方法。首先，对采集的刀具图像进行高斯滤波获得高斯滤波图，消除噪声信息；然后，通过高斯差分滤波获得高斯差分图，提取刀具背景纹理信息；在此基础上，利用高斯滤波图与高斯差分图之间的中央-周边操作获得显著图，以消除高频噪声信息以及光照不均等低频背景信息，提高刀具磨损区域的显著性；最后，根据刀具磨损区域特征进行刀具磨损区域分割和滤波。实验表明，推荐方法能够准确检测刀具磨损区域，具有较高的检测准确率。<br>In order to improve the accuracy of tool wear area detection, a new method for tool wear detection based on the characteristics of tool wear area is proposed. Firstly, the tool image is filtered by using Gauss filter to eliminate the noise information and gain a filter image. Then, the background texture information is extracted by using the difference of Gaussian filter to get a difference image. On the basis, the center-surround operation is used to obtain the saliency map between the filter image and the difference image, by which can eliminate high frequency noise and uneven illumination of low frequency background information, and improve saliency of the tool wear area. Finally, the tool wear region is segmented and filtered according to the characteristics of tool wear area. The experimental results show that the method can accurately detect the tool wear area and has higher detection accuracy
%K 刀具磨损
%K 高斯差分滤波
%K 中央-周边操作
%K 显著图&lt
%K br&gt
%K tool wear
%K difference of Gaussian filter
%K center-surround operation
%K saliency map
%U http://journals.nwpu.edu.cn/jxkxyjs/CN/abstract/abstract6946.shtml