摘要
从零件图像的小波分解系数和图像相对边缘像素系数作为零件特征的方法出发,提出了基本概率分配构造和多源零件图像特征识别的方法。首先,对多源零件图像分别进行小波分解和多尺度边缘检测,获取零件图像的小波分解系数和相对边缘像素系数。然后,将它们作为神经网络的输入,获取多源零件图像识别的基本概率分配。最后,依据证据理论的合成规则得到零件的识别结果。实验结果表明,所提出的方法是有效的。
It uses the coefficients of wavelet transform and the relative edge pixel coefficients of image to represent the part features, presents a method for basic probability assignment contribution and feature recognition of multi - source part image. By analysis of the multi - source part image and detecting the edges with wavelet transform, it obtains the coefficients of wavelet transform and the relative edge pixel coefficients, these can take as the inputs of a neural network to obtain the basic probability assignment. The part is realized pattern recognition using combination rules of Dempster- Shafer evidential theory. The experiment results show that the proposed method is effective for feature recognition of multi - source part image.
关键词
小波变换
神经网络
证据理论
图像识别
Wavelet Transform
Neural Network
Evidential Theory
Image Recognition
