Object recognition in household environment using fusion of shape and Haar-like features

In this paper we present some preliminary results of our project concerning about developing an image recognition technique for detection of certain household objects, mainly based on the shape and the Haar-like features. Shape feature is used to find the position of certain shape household objects preliminarily. Then, Haar-like feature-based recognition may be only performed for small regions around the positions of detected shape features. Thus, the proposed two-stage recognition can reduce the computation. In addition, experimental results are also presented to prove that the fusion of the two feature detection strategy will complement each other and improve the detection rate and the robustness.

[1]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[2]  Tomaso A. Poggio,et al.  A general framework for object detection , 1998, Sixth International Conference on Computer Vision (IEEE Cat. No.98CH36271).

[3]  Erkki Oja,et al.  A new curve detection method: Randomized Hough transform (RHT) , 1990, Pattern Recognit. Lett..

[4]  R. A. McLaughlin,et al.  Randomized Hough transform: better ellipse detection , 1996, Proceedings of Digital Processing Applications (TENCON '96).

[5]  Andrew W. Fitzgibbon,et al.  Ellipse-specific direct least-square fitting , 1996, Proceedings of 3rd IEEE International Conference on Image Processing.