2D partially occluded object recognition using curve moments

This paper presents a new approach for recognition of 2D partially occluded objects using the curve moments as the features. In this approach, the boundary of object of interest is first extracted after image pre-processing. Then corner points were used to partition the boundary into curve segments consisted of 3 consecutive breakpoints. Subsequently, seven different order moment descriptors are computed as feature vectors for each segment. Finally, feature matching between the object of interest in scene and the model is performed hierarchically. From the experiment results, the proposed recognition algorithm was found to be robust to similarity transform, noise and partial occlusion, and computational efficient.