Shadow Detection – Its Pros and Cons

Location Based Service Domain offers various services to applications like Navigation, Traffic Surveillance etc. One of the major services of this domain is the tracking moving objects. This service is implemented as system design approach having inputs as photos or videos. These inputs have many entities like the object of interest and entities of confusion .One such entity of our interest which may lead to major failure of the system is the shadow of the objects of interest which may be mistook as the object itself . This will lead to false detection. Thus shadow detection and its elimination is proposed as a necessary and important step of the system. For this the objects have to be classified as Static and dynamic with shadow regions as umbra and penumbra. For the design of the proposed preprocessing stage, the paper evaluates different types of shadows, Application wise, Model wise, Method wise and Property wise. Finally for the analysis phase of the system, a comparative study of some required parameters for shadow detection is done. The system is then evaluated to conclude the advantages and disadvantages of shadow detection preprocessing module General Terms This paper is in domain of image processing. The sub domain being Tracking in which the shadows of the tracked objects are detected and eliminated.