论文部分内容阅读
以计算机视觉自动检测果实表面着色度并进行分级为目的，建立了室内计算机视觉系统获取苹果果实的彩色图像，并将ＲＧＢ值转换成ＨＬＳ值；在分析苹果颜色特性的基础上，确定了用合适色相值下累计着色面积百分比进行颜色分级的方法。分级试验结果表明，用建立的准则和方法，计算机视觉分级与人工分级的一致度在８８％以上。
In order to automatically detect the degree of coloring and classify the fruit surface by computer vision, an indoor computer vision system was established to obtain the color image of apple fruit and convert RGB value to HLS value. Based on the analysis of apple color characteristics, The method of color grading based on the cumulative coloring area percentage under hue value. The grading test results show that the consistency between computer vision grading and manual grading is above 88% with the established guidelines and methods.