Spatial Verification for Scalable Mobile Image Retrieval

Abstract

Owing to the portable and excellent phone camera, people now prefer to take photos and upload them by mobile phone. Content based image retrieval is effective for users to obtain relevant information about a photo. Taking the limited bandwidth and instability into account, we propose an effective scalable mobile image retrieval approach in this paper. The proposed mobile image retrieval algorithm first determines the relevant photos according to visual similarity in mobile end, then mines salient visual words by exploring saliency from multiple relevant images, and finally we determine the contribution order of salient visual words for scalable retrieval. In server terminal, spatial verification is performed to re-rank the results. Compared to the existing approaches of mobile image retrieval, our approach transmits less data and reduces the computational cost of spatial verification. Most importantly, when the bandwidth is limited, we can transmit a part of features according their contributions to retrieval. Experimental results show the effectiveness of the proposed approach.

From IR to Mobile IR

“…Whence, there is the need of new and specific retrieval systems for "mobile images", coping both with networking and hardware limitations of mobile devices. For instance, Yang and Qian [200] propose a novel approach that leverages on the fact that often users take multiple shots of a given scene; hence, the mobile device is first searched for photos visually similar to the query image. Then, the latter, together with the relevant photos found in the device, is used to mine "visual salient words", which are ordered according to their contribution in order to reduce the noise and the computational complexity of spatial verification.…”

Improved image GPS location estimation by mining salient features

“…However, feature matching based approaches is both low accuracy and has a heavy computational cost for a large scale dataset. To solve the problem, bag-of-word (BoW) model based approaches are proposed [31][32][33][34][35][36][37][38]. In these approaches, fast image retrieval is achieved by giving each BoW a weight and then computing scores for each image in the dataset.…”