A novel method for discovery of objects

We present a novel method for detecting and removing shadow from unprocessed raw images without any additional user input. Shadow detection and removal is an important pre-processing step in several applications. All previous method proposes different algorithm to detect shadow from images, but they suffer poor performance while performing distinction of dark objects with shadow and its removal from input images. Proposed algorithm is capable of detecting and removing shadow based on a concrete detection algorithm and a light model with two type of light. Shadow detection is performed by using input and near infrared image (NIR) of input. Removal is done by the help of a binary mask derived from the input image. The proposed method can achieve detection and removal of shadows with a time complexity of 1.06 seconds. Experiments shows our method achieves a better shadow detection and removal than other existing methods.