摘要
利用计算机视觉技术提取稻谷的主要外形特征,初步建立了稻谷品种数据库。研究了稻谷品种数据库的检索技术,提出了将BP神经网络与相似度检索相结合的检索方法。实验结果表明,该方法可有效提高检索效率。
A database was established to store external features of rice varieties extracted by computer vision. Retrieval technique of rice image based on content was investigated and a new retrieval method was proposed by combining the BP neural network with the similarity retrieval technique. The tests showed that the new method is of high efficiency both in identification and in database retrieval of rice variety.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2005年第10期94-96,共3页
Transactions of the Chinese Society for Agricultural Machinery
基金
教育部留学回国人员科研启动基金资助项目(项目编号:200112)
关键词
稻谷
数据库
检索
计算机视觉
Rice, Database, Retrieval, Computer vision