计算机视觉技术在番茄损伤区域自动检测中的应用

Study on Automatic Identification of Tomato with Bruise Based on Computer Vision

农产品品质检测与分级是流通和加工过程中的一个重要环节,其检测与分级水平不仅直接关系到人民的健康,而且还是国际、国内农产品市场竞争的重要因素。为此,研究了图像处理的去除噪声、图像分割、图像增强等多种低层处理的方法,建立用区域增长法进行番茄表面缺陷区域检测,实验证明番茄损伤检测的准确率不低于90%。

Three methods that include filtering noise and dividing of image and highlighting of image etc. for lower-layer image processing technique were studied in this paper, and Distriction increasing for detecting bruise image of tomato was built. The experiment results showed that the rate of testing precision was not less than 90%.