提出基于LSH索引的遥感图像检索技术,结合离线建立索引的思想,实现了遥感图像快速检索.并借鉴特征选择的思路,提出了一个新的索引有效性指标——基于特征鉴别能力的索引有效性指数(feature discriminative-ness-based indexing validation index,FDIVI),通过对各个特征空间上的LSH索引评估选择最能区分目标和背景的特征.对若干典型地表覆盖的遥感图像检索实验证明了LSH索引及FDIVI指标在遥感图像检索系统中的有效性. A remote sensing image retrieval technology based on LSH index is proposed, and the idea of ​​offline indexing is developed to realize rapid retrieval of remote sensing images. A new index validity index is proposed based on the idea of ​​feature selection - index validity based on feature identification ability Index discriminating index (FDIVI), which distinguishes the target from the background by selecting and evaluating the LSH index of each feature space.Experimental results of several typical land cover remote sensing images show that the LSH index and The Effectiveness of FDIVI Indicators in Remote Sensing Image Retrieval System.