Method for generating depth maps from monocular images and systems using the same

摘要: Methods for generating depth maps from monocular still image or video and systems using the same are provided. First, an initial map is estimated arbitrary defined. For inputs, motion information can be used, background set by default, chosen user estimated. Estimation of based on a contrast blur map. The defines values respective pixels picture frames. original frame data mapped to according positions pixels, in order obtain corresponding values. An space subdivided into plurality sub-image spaces, value each filtered located space, pixels.