摘要：【中文】本发明公开了一种根据特征提取算法定位着丝粒的图像处理方法，本发明根据每一号的每一条染色体的着丝粒位置作为样本数据提取着丝粒的范围；然后统计好每号染色体的范围选择后再根据每号染色体通过特征提取、去噪、图像预处理、图像分割等技术处理染色体图像来准确的定位到着丝粒的范围，求范围内的每条染色体的法线并把最短法线所在位置为着丝粒所在的位置。使得在测试的1号至22号染色体和X染色体的准确率达到99%以上，Y染色体的准确率达到98%以上，每号染色体的数据至少为500。 【EN】The invention discloses an image processing method for positioning centromere according to a feature extraction algorithm, which is used for extracting the range of the centromere according to the position of the centromere of each chromosome of each number as sample data; then counting the range selection of each chromosome, processing chromosome images through the technologies of feature extraction, denoising, image preprocessing, image segmentation and the like according to each chromosome to accurately position the chromosome images to the range of the centromere, solving the normal of each chromosome in the range and setting the position of the shortest normal as the position of the centromere. The accuracy rate of the tested chromosomes 1 to 22 and the X chromosome is more than 99 percent, the accuracy rate of the tested chromosome Y is more than 98 percent, and the data of each chromosome is at least 500.