Partial View-Based Object Recognition using Leave-one-out Approach with Classification and Regression Trees

S.M. Mohidul Islam; Farhana Tazmim Pinki

Abstract

Object recognition from image is an intriguing task in computer vision. It is simple for human being but complex for machine because of complex background, lightening, different viewing angle, size etc. So, a typically automatic and robust system is required to recognize object. In this paper, a content-based object recognition system using only one view of the object: top view, front view, back view, side view, and also using combination of these views to recognize object from image is proposed. Color moments as well as wavelet packet entropy, which are invariant to translation, rotation, and scaling are used as features to recognize one object from another using Classification and Regression Trees. The validation method, Leave-one-out approach is followed to train and test the model. Experimental results are performed using the publicly available Cogvis ETH-80 dataset. The experimental results show that this system can perform comparable recognition of object using partial view of the object and the evaluation results conclude that the proposed system shows better results than many existing methods.Cite this ArticleMohidul Islam SM, Farhana Tazmim Pinki, Partial View-Based Object Recognition using Leave-one-out Approach with Classification and Regression Trees, 2018. Journal of Computer Technology & Applications. 2018; 9(3): 9–16p.

Abstract

Keywords