This thesis proposes a robust shadow detection method for single input natural image followed by a shadow removal method for adequate reconstruction of shadow obstructed path edges. The proposed shadow detection method emplys spectral rationing and automatic thresholding techniques in detecting shadow regions. The extracted shadow regions are then processed for shadow removal. As adaptive illumination transfer shadow removal method is presented to identify and reconstruct the shadow obstructed path edges. Experimental results depicts the effectivess of the proposed shadow detection and removal algorithm on shadow obstructed path edges has resulted an acceptable path edge construction.