DOI: 10.3724/SP.J.1187.2013.00647

Journal of Electronic Measurement and Instrument (电子测量与仪器学报) 2013/27:7 PP.647-652

Object recognition method based on salient regions and BOW model

Aiming at the problem that many local features are unstable, unreliable or irrelevant in constructing visual dictionaries, a new object recognition method based on salient regions and BOW model is proposed. Firstly, this method utilizes the strong corner detector to determine the saliency of the image. Then, the local features are extracted from the saliency of the image and modeled as BOW. The final results are obtained by the nearest neighbor classifier. The proposed method is evaluated on different datasets which varies largely in object scale, object viewpoint and background condition. The experimental results show that the proposed method outperforms than existing object recognition methods both in recognition rate and robustness, the average recognition rate can reach 96.13%, 82.06% and 66.7%.

Key words:BOW,strong corner detector,saliency,object recognition

ReleaseDate:2014-07-21 17:01:58