OBJECT RECOGNITION BY A LINEAR WEIGHT CLASSIFIER

Most of the current object recognition systems involve matching the sc ene object against every predefined object in the model database one a t a time. The major problem with this approach is that as the number o f models is increased the computational complexity and time requiremen ts of the systems are greatly increased. In this paper, we present a l inear supervised classifer that uses a weight vector as a filter to mo dify the now of the input information. The optimal weight vector is de rived from a set of training samples in the sense of minimum least-sqa res error between desired and actual output responses. The computation al complexity of the weight vector is independent of the number of mod els. Assignment of a scene object to one of several classes of objects is directly determined by the weight vector and the input image of th e object. Experimental results on both man-made workparts with simple shapes and natural objects with complex shapes are studied in this pap er. The occluded and distorted versions of these test objects are also included to evaluate the performance of the linear weight classifier.