基于内容的图像检索技术在农业信息化中的应用

The Technique of Content-based Image Retrieval and the Applicationin Agriculture

在基于内容的图像检索技术的基础上,针对农作物图像数据库的特点,提出了一种综合使用颜色、纹理和形状特征的新的检索算法,使检索的精确度得到了很大提高,并且通过合理设置特征库,大大提高了检索的效率。为此,通过编程构造了一个实验系统,组织了一个相对完备的农作物图像测试数据库和通用图像库,验证了新的检索算法的有效性,并对利用不同特征进行检索的性能进行了评测和比较,同时对出现的问题进行了分析。

Based on the study of content-based image retrieval techniques,this paper puts forward a novel retrieval algorithm synthesized using color,texture and shape features aimed for crop picture database which can greatly improve the accuracy to the retrieval,and proposes a reasonable image feature database through which the retrieval efficiency can be improved significantly.A retrieval system is programmed in the article,and a crop image database and a general image database is organized to test the validity of the new method.This paper tests and compares the efficiency of the different character based retrieval methods,and analyses the problem emerged.