科研之友
微信
新浪微博
Facebook
分享链接
摘要
Computer vision can greatly benefit from the development of image intensity model. The interpretation of image intensity requires a sound understanding of the various mechanisms involved in the imaging process. It is also essential to research and utilize image intensity model for developing and refining algorithms to extract information from images. In this paper, image intensity model in the daylight based on geometrical optics is studied by utilizing Lambert model. Firstly, the digital imaging process and the optical analysis methods are introduced. Secondly, the daylight is decomposed into the sunlight and the skylight, and the solar illumination and the skylight illumination incident on the surface of object are analyzed respectively. Thirdly, the solar irradiance and the solar flux, the skylight irradiance and the skylight flux incident on the object surface element are given respectively, where the main steps in the derivations of the two illumination models are described. Then, the irradiance incident on the imaging plane surface element and the image intensity model are derived. This image intensity model is comprised of five primary components, the imaging system gain term, the object surface property term, the object space state term, the object surface illumination term, and the intensity background noise term. The image intensity model indicates that image intensity is dependent on the illumination incident on object surface, the objects surface property, and the object surface attitude relative to the imaging plane and it is influenced by the image intensity background noise of the imaging system.
