An Adaptive Tracking for Moving Targets in Shadows and Poor Illuminations

Abstract

Tracking is an interesting area of research. It has to meet several challenges in real-time. There is a more noteworthy possibility of missing the objective. Again, the target may be single or multiple. Each one is having its speed. The biggest obstacle is when the target meets the shadow of some object or itself, the intensity of the target changes severely. This results in missing the target. In this proposed paper, an adaptive algorithm using the Difference method with normalized values indexed with Vegetation parameters (NDVI) is utilized to differentiate the target from shadows followed by tracking the desired object which may be moving at different speeds by an improved optical flow algorithm. Thus, this proposed method aims to follow the object along with challenging illuminations and in the presence of shadows. This is particularly useful to follow the objects successfully on both internal and out-of-doors scenes in shadows. Examination and comparison of various videos in MAT Lab with standard data set indicate that this method yields a better tracking result with state of art methods.