一种基于RGB颜色分量的枝干图像分割方法

摘要

On the basis of the RGB color space, aiming at the recognition of fruit tree branches, a method for threshold segmenta⁃tion of color image applied to the recognition of fruit tree branches is proposed. By analyzing the relationship between every two components of the three base color components RGB, the linear relationship between every two components of the three base color components is obtained, and then the concept of chromatic aberration ring is put forward, and the chromatic aberration ring of the images of the fruit tree branches is set up, the connection between the linear expressions of the base color components is obtained. Furthermore, the image segmentation threshold range formula is drawn. A parameter μ to adjust threshold is introduced, which makes it possible to adjust the threshold value according to the real condition of the orchard in real⁃time. The experimental verifica⁃tion shows that this method can execute the fast segmentation and recognition of the branches in the color images of fruit tree bran⁃ches under the natural light.%在RGB色彩空间的基础上，以苹果枝干识别为目标，提出了一种用于枝干识别的彩色图像阈值分割的方法。通过分析RGB三个色彩分量的关系，得出了基色两两之间的线性关系表达式，进而提出色差环的概念，并建立了色差环，取得线性关系表达式之间的联系，得出枝干图像分割阈值范围公式。引入对阈值进行整体调节的参数μ，实现根据实际果园环境实时调节阈值范围大小。通过实验验证，本文方法能够实现对自然光照射下的果树枝干等彩色图像的准确快速分割和识别。