水果品质机器视觉检测中的图像颜色变换模型

在水果品质的机器视觉检测过程中，水果形状导致的图像颜色失真直接影响最终的检测和分级结果。本文在6个位置分别采集了3种大小、19种不同表面颜色的标准实验球体的图像。在分析球体的大小、表面颜色和图像采集位置等因素对球体图像颜色的亮度、色调和饱和度的影响后发现，球体图像的颜色失真主要是颜色的亮度失真。建立了摄像视区中心球体图像的亮度校正模型，该模型的相关系数R^2为0．846。采用该模型在Matlab软件平台上构建了柑桔图像的颜色校正算法，实验结果表明，该颜色校正模型和算法是有效和可靠的。

Color Transformation Model of Fruit Image in Process of Non-destructive Quality Inspection Based on Machine Vision

In the process of fruit quality inspection, color distortion of fruit image resulted from the three dimensional shape of fruit directly influences the accuracy of inspection and classification. In order to reduce this color distortion, the images of 57 balls, which were divided into three groups with diameters of 75 mm, 60 mm, and 40 mm respectively and each group consisted of 19 balls with different colors, were captured at 6 different positions. After analyzing the influences of size, color and image capturing position of balls to brightness, hue and saturation of ball image, it was concluded that the color distortion of ball image consisted mostly in its brightness distortion, and the model of image brightness correction at the center of camera view field was developed, with a correlation coefficient R 2 of 0 846. Based on this model, the color correction algorithm of citrus image was put forward, and the experimental results verified that the color correction model and algorithm were efficient and reliable.