基于RGB颜色模型的红富士苹果表皮红色区域检测

Detection of Red Region of Fuji Apple Based on RGB Color Model

水果的颜色是水果分级的重要依据,影响着消费者的购买欲望。研究了一种基于RGB颜色模型的红富士苹果表皮红色区域检测算法。通过苹果RGB图像的光照补偿,降低光照变化和不均匀性带来的影响。在此基础上,计算各像素R、G、B分量的G/B和R/G比值,通过训练获得其分割阈值,实现苹果图像和背景的准确分割。最后用超红-超绿阈值分割法检测分割后的苹果图像的红色区域并计算其面积。实验结果表明:基于RGB颜色空间的红富士苹果表皮红色区域检测算法能够准确地检测出果皮表面的红色区域,满足红富士苹果颜色等级检测的需要。

Color of fruits is an important index for the classification of fruits, which will affect consumers′ purchase desire. An algorithm based on RGB color model is studied to detect the red area of Fuji apple. The light compensation of image is used to reduce the effect of illumination change and nonuniformity of light source. Two parameters of R/G and G/B in RGB image are calculated, and the thresholds of these two parameters are obtained by using training samples to achieve accurate segmentation for apple and background. The excess red minus excess green threshold segmentation method is applied to detecting and calculating the red region of apple surface. The experimental results show that the algorithm based on RGB color model is able to accurately detect the red area of the apple surface. The algorithm can meet the requirement of color grading detection of Fuji apples.