Zahra Hossein-Nejad - Department of Electrical Engineering, Sirjan Branch Islamic Azad University Sirjan, Iran
Mehdi Nasri - Department of Biomedical Engineering, Khomeinishahr Branch Islamic Azad University Isfahan, Iran

Object Recognition System is widely used in different real-life applications such as content-based image retrieval, object detection, etc. In this article, we suggest a noveltechnique for object detection using Redundant Keypoint Elimination method SIFT- Graph Transformation Matching (RKEMSIFT-GTM). This proposed approach deletes redundant points and eliminates false matches. The proposed improved region-growing, which is a powerful method, is used for the final detection stages. The suggested approach is evaluated on datasets such as COIL-۱۰۰ and obtained a good recognition rate compared to other detection methods.

Object Recognition; matching; SIFT; GTM.