为了准确识别并分割套袋苹果,对影像中的套袋苹果及其背景在RGB颜色空间模式下,分析各目标物的R(红)、G(绿)及B(蓝)值分布特征,根据各目标物所呈现的R、G及B值差异选取适合于套袋苹果的分割条件,对套袋苹果进行分割,将分割结果与多阈值分割法以及K均值聚类法分割结果进行比较。结果表明:分割法优于以上2种方法,同时,根据该试验提出的分割方法计算苹果识别率,识别率可达93.4%,表明该试验算法对套袋苹果的分割准确率高。 In order to accurately identify and segment bagged apples, the distribution characteristics of R (red), G (green) and B (blue) values ​​of each target were analyzed under the RGB color space mode for images of bagged apples and their backgrounds. The difference of R, G and B presented by each target selects the suitable segmentation conditions for the bagged apple, and divides the bagged apple, and compares the segmentation result with the multi-threshold segmentation method and K-means clustering method. The results show that the segmentation method is superior to the above two methods. At the same time, the recognition rate of apple is calculated according to the segmentation method proposed in this experiment, the recognition rate can reach 93.4%, indicating that this algorithm has high segmentation accuracy for bagged apple.