Support Vector Machine for classify dynamic human/vehicle shapes

Currently support vector machines (SVM) became subject of interest because of its ability to give high classification performance in a wide area of application. Most of the classifier model especially based on supervised learning involve complicated learning model and yet the performance sometimes worst. This paper proposes a SVM model to classify between human and vehicle shapes in various pose. SVM classify data by first construct a decision surface that maximizes the margin between the data. For testing new data, SVM will calculate the sign signifying where this new data reside in the constructed decision surface. The developed model will be used to classify an outdoor scene of human and vehicle shapes in dynamic pose. Results of the experiments showed a satisfied performance with the proposed approach.