本文展开了对尿沉渣图像

自动识别算法的研究。著者在从事“尿沉渣图像自动识别算法的研究”课题的研究

中，广泛吸取了国内外已有的有益成果，根据尿沉渣图像特点从图像增强、图像

分割、特征提取和特征选择、图像识别等各个环节寻求最佳的可行方法以提高识

别的性能，提出了基于组合图像分割算法、改进的特征提取和特征选择算法以及

BP神经网络分类器的尿沉渣图像自动识别算法。-This started on the urine sediment image recognition algorithm. Author engaged in " Urine image automatic recognition algorithm of the project on"　study, draw extensively on the achievements of the existing home and abroad, according to the image characteristics of urinary sediment from the image enhancement, image segmentation, feature extraction and feature selection, image recognition each link to find the best possible ways to improve the recognition performance of this composite image segmentation algorithm based on improved feature extraction and feature selection algorithm and BP neural network classification of devices in the urinary sediment Tuxiangzidong recognition algorithm.