Feature based three-dimensional object recognition using disparity maps

The human vision system is able to recognize objects it has seen before even if the particular orientation of the object being viewed was not specifically seen before. This is due to the adaptability of the cognitive abilities of the human brain to categorize objects by different features. The features and experience used in the human recognition system are also applicable to a computer recognition system. The recognition of three-dimensional objects has been a popular area in computer vision research in recent years, as computer and machine vision is becoming more abundant in areas such as surveillance and product inspection. The purpose of this study is to explore and develop an adaptive computer vision based recognition system which can recognize 3D information of an object from a limited amount of training data in the form of disparity maps. Using this system, it should be possible to recognize an object in many different orientations, even if the specific orientation had not been seen before, as well as distinguish between different objects.