Abstract: Using machine vision technology to quickly analyze the agronomic traits of pepper can provide a basis for pepper genetic breeding and functional genomic analysis. Based on Labview, Visual Studio and OpenCV software, a set of analysis programs for agronomic traits of pepper cross-cutting and longitudinal cutting sections were written in this paper. The cross-section images of pepper were analyzed and processed, and the outer contour, inner contour, ventricular map, wrinkle graph and other images were drawn on the cross-section. The longitudinal section was marked with operations such as the longitudinal section skeleton of pepper, the end point of pulp, the middle line of fruit stalk and so on. Analyzing the images and extracting data, the parameters of the transverse section were extracted such as diameter Dx and Dy, circularity DC, pulp thickness T, ventricular number H, transverse folds VWN and VWD, and the parameters of the longitudinal section were extracted such as mid-long line Y, shoulder width X1, chest width X2, abdominal width X3, tip width X4, midline path R. The automatic and high-throughput extraction of the agronomic characters of the hot pepper slices was realized.

Key words: Capsicum annuum L., image processing, crosscutting, slitting

摘要： 利用机器视觉技术快速分析研究辣椒的农艺性状，可以为辣椒选育提供依据，后续为辣椒遗传育种和功能基因组分析打下坚实的基础。基于Labview、Visual Studio和OpenCV软件编写了一套对辣椒横切和纵切截面农艺性状的分析程序。对辣椒截面图像进行分析处理，对横截面提取外轮廓、内轮廓、心室图、褶皱图形等图像，对纵截面提取辣椒骨架、果肉端点、果柄中线等操作。对这些图像进行分析和数据提取，提取出横截面的直径Dx和Dy、圆形度DC、果肉厚度T、心室个数H、横向褶皱VWN/VWD等参数，提取纵截面的中长线Y、肩宽X1、胸宽X2、腹宽X3、尖宽X4、中线形状等参数。实现了对辣椒截面农艺性状的自动化、高通量提取。

关键词: 辣椒, 图像处理, 横切, 纵切