International Conference on Instrumentation, Measurement, Circuits and Systems (ICIMCS 2011)
181 Near-Duplicate Image Detection Based on LSH
Download citation file:
- Ris (Zotero)
- Reference Manager
Detecting near-duplicate(ND) images efficiently and accurately is very important for copyright infringement detection and redundancy removal in retrieval results. In order to solve the low efficiency of existing methods, a near-duplicate image detection method based on Exact Euclidean Locality Sensitive Hashing(LSH) is proposed. Firstly, LSH is used to filter most of the non-near-duplicate image pairs. Then, a pool of vastly reduced image pairs is further explored with Enhanced Scale-Rotation invariant Pattern Entropy(ESR-PE) to remove the wrong matches, thus near-duplicate image detection is accomplished. Experimental results show that the novel method significantly speed up the detection with competitive performance to the exhaustive detection.