基于BP网络的棉花中异性纤维识别算法

针对原棉中的异性纤维种类繁多、特征各异,很难建立一个统一的异性纤维识别模型问题.提出了应用神经网络识别异性纤维的算法.在RGB颜色空间下,提取原棉图像中每个像素的R、G、B分量,采用统计纹理分析方法,计算R、G、B分量的纹理强度信息作为特征值.用MATLAB软件创建BP网络识别模型,对有代表性的原棉图像进行识别实验.实验结果表明该算法可以有效地识别出原棉中存在的有色和无色异性纤维.

Recognition algorithm of the foreign fibers in cotton by BP neural networks

Owing to numerous kinds of foreign fibers and varied characteristics in cotton,it is difficult to establish a universal recognition model of foreign fibers.In order to solve this problem,recognition algorithm of the foreign fibers in cotton based on BP Networks is proposed.In RGB color space,R,G and B values of each pixel in cotton image are extracted and used to calculate texture intensity values about R,G and B as eigenvalues based on analysis method of statistical texture.MATLAB software is used to set up BP recognition model and test some typical cotton images.The results show that the algorithm can recognise the colored and uncolored foreign fibers in cotton effectively.