摘要
通过对芒果成熟期3种品质指标测定的同时,拍摄数字相片,以计算机图像分析技术为手段,采集芒果颜色参数。研究3种颜色模型中与芒果品质变化相关的颜色特征量,归纳出合适的颜色参数与芒果品质指标值间关系式,用于推导芒果的成熟度。并结合形状、表面斑痕、颜色分布等特征,经图像处理和运算,抽取出所需的几何尺寸和突变部位的数据,作为样品分级的依据。以期建立一种与人工目视辨别有相同精确度、高效率、客观的自动选别评价体系。
Some maturity indicators of mango were detected and digital images were taken at the same time during storage period. Using the computer image analysis technology, the color parameters of mango surface that related to quality indicators were studied, and the regression equations were concluded to forecast mango＇s quality. Moreover, other parameters of the shape, size and splash surface were processed with image analysis for providing evidences of classification. This study will be helpful for building an efficient, accurate and objective mango automatic sorting system which is comparable to the human eyes.
出处
《湖北农业科学》
北大核心
2009年第8期1992-1995,共4页
Hubei Agricultural Sciences
关键词
芒果
图像处理
特征抽取
自动分级
mango
image processing
characters
automatic sorting
