An Adaptive Classifier for 3-D Planar Object Recognition Based on Uncertainty of Features by Binocular Stereo Method

Binocular Stereo 방법에 의한 3차원 평면 물체의 특징값의 불확실성을 고려한 적응분류기

In this paper, we propose an adaptive classifier based on uncertainty of features for 3D planar object recognition. First, we investigate the uncertainty of depth information and the feature values of 3D planar object by numerical method. And, we observed that the statistical behavior of feature is dependent on the position and orientation of objects. After that, the approximation of the statistical behavior is executed. Subsequently, the recognition procedure is executed by the adaptive classifier. By computer simulation, we confirmed that the proposed classifier is useful for 3D planar object recognition.