Abstract | ||
---|---|---|
State-of-the-art near-duplicate image retrieval systems take the image as a whole by the bag-of-words (BOW) representation. Feature quantization on large image database always reduces the discriminative power of image features, and the global BOW feature neglects the geometric relationships among local features. We propose in this paper a region-based image retrieval method. Image similarity is determined by the consistent MSER regions of two images, while the matched MSER regions is identified by similar local features (e.g. SIFT), where both the feature vectors and the geometric consistency among the feature locations are considered to ensure stable matching. Moreover, all the operations are realized on the original features but not the quantized ones to avoid the loss of the discriminative power. Experimental results confirm the superiority of the proposed method, especially the retrieval precision at top ranked images. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1109/ACPR.2013.131 | ACPR |
Keywords | Field | DocType |
feature quantization,near-duplication image retrieval,global bow feature,image representation,image similarity,local features,image matching,feature location,feature vector,discriminative power,visual databases,region-based near-duplicate image retrieval,state-of-the-art near-duplicate image retrieval,image features discriminative power,bow representation,image mser region matching,image feature,top ranked images,bag-of-words representation,feature extraction,image retrieval,region-based matching,region-based image retrieval method,large image database,feature vectors,mser,geometric consistency,image database | Computer vision,Feature vector,Automatic image annotation,Feature detection (computer vision),Pattern recognition,Image texture,Feature (computer vision),Image retrieval,Feature extraction,Artificial intelligence,Mathematics,Visual Word | Conference |
Citations | PageRank | References |
0 | 0.34 | 8 |
Authors | ||
4 |