干扰图像快速检索算法

Interference Image Quick Search Algorithm

文章提出了一种适用于大规模数据集的图像检索算法，目的是在有干扰的情况下，准确地找出目标图片。本文采用局部特征算子作为衡量标准，又创造性地在每个视觉词袋后[1]建立图像索引，形成以特征算子为基元的带权视觉词典结构，在保证检索精度的前提下尽可能的提高检索速度，更好的满足了大规模数据对效率的要求。实验证明，该算法具对存在一定干扰的图像效果显著，鲁棒性强，达到了国际一流水准。

This paper presents a large-scale data sets for image retrieval algorithm,the aim is the case of interference,and accurately identify the target picture.In this paper, the local operator as a measure of standard features,but the visual creativity of each word in the bag[1]to establish an image index,to form a characteristic element of the operator-based visual dictionary with the right structure to ensure the accuracy of the premise to make retrieval possible to improve the retrieval speed,and better meet the requirements of large-scale data on the efficiency.Experiments show that the algorithm with the image of the interference .effect of certain significant,robust,reaching a world-class standards.