基于机器视觉技术的西瓜瓜瓤的识别方法

Identification of the Machine Vision Technology Based on Watermelon Flesh

为了实现西瓜瓜瓤和瓜皮的自动识别,提出了利用计算机视觉技术代替人眼对瓜皮和瓜瓤进行自动识别。采用Matlab软件平台构建算法。首先对西瓜沿中心线切开,用CCD摄像机对西瓜剖面图像进行采集,通过红绿色差法对剖面图像进行阀值分割,实现瓜瓤从瓜皮中分离,利用类椭圆确定瓜瓤的圆心、长轴和短轴。实验证明,该算法和程序能够识别瓜皮和瓜瓤,以及对瓜瓤的相关计算。

In order to realize the automatic recognition flesh and rind of watermelon,computer vision technology is put forward to replace human eyes for the recognition rind and pulp. The construction algorithm is using Matlab software platform. Firstly,along the center line incision,watermelon profile image is collected by CCD camera,and the profile image threshold segmentation by red green color difference method,to achieve flesh from the rind,and use the elliptic determine the center,major and minor axes of flesh. Experiments show that,the algorithm and the program can identify the rind and pulp,and related calculation of pulp.