An efficient sky detection algorithm based on hybrid probability model

Sky is one of the most significant subject matter commonly seen in outdoor photos. We propose a highly efficient sky detection algorithm. First, we detect a rough sky-ground boundary. Then, we calculate the parameters related to appearance of sky. Finally, we use these parameters to construct a hybrid probability model that indicates how possible a pixel belongs to sky. Moreover, an image processing library with parallel processing techniques is embedded in the proposed algorithm. The proposed algorithm has both high accuracy and high efficiency and can process a video image in real time. If the input is a VGA size image, the computation time of the proposed algorithm is less than 35ms when using a common desktop.