摘要
充分利用计算机视觉和人工神经网络技术 ,建立了一个苹果综合外观品质检测与分级系统 ,实现了对苹果的正确分级 ,正确识别率可达 90 .8%。还借助于遗传算法对 BP网络进行设计 ,提高了其学习速度。
Computer vision and artificial neural network techniques were used to establish an integrated detecting and grading system for external qualities of apple.The results of the experiment showed that the system can correctly grade apples and achieve accuracy of 90.8%.A GA-based training algorithm was introduced to find optimal structures or the number of hidden layer nodes and connection weights of the two neural networks,and the speed of training was found improved.
出处
《西北农林科技大学学报（自然科学版）》
CSCD
北大核心
2001年第6期108-111,共4页
Journal of Northwest A&F University(Natural Science Edition)
基金
国家自然科学基金资助项目 ( 4 130 10 3)
