摘要
水果分级与表面缺陷检测,是水果自动分级检测系统的基础。提出表征水果类别的"大小-评价测光值"空间模型,与表征水果表面缺陷程度的"数量-程度"空间模型。将水果的特征映射到这两个特征空间,然后进行模式分类。实验结果表明,提出的两个特征空间能有效的表征水果的特征,准确地划分水果的类别与表面缺陷程度。
Fruit grading and surface defect detection, is the foundation of the fruit automatic classification system. Size-evaluation metering space as the fruit categories model, Volume - Degree space as the fruit surface defect model are presented. Fruit' s characters will be mapped to the two space and then proceed pattern classification. Experiments show that the two spaces can be effective characterization fruit' s characteristics, classification its type and degree of surface defects accurately.
出处
《计算机工程与设计》
CSCD
北大核心
2008年第15期3954-3957,共4页
Computer Engineering and Design
基金
云南省自然科学基金项目(2005F0012M)
关键词
图像处理
模式识别
生物目标识别
水果检测
水果分级
缺陷检测
image processing
pattern recognition
biological target identification
fruit detection
fruit classification
defect detection