A new method of simulating focusing gradient based on fast stratified filtering on 3D image

The relationship between fuzzy function a and the distance from object to lens and focal Length was provided by using the lens model. A new adaptive stratified filtering method for 3-D depth image was developed. The concept of the difference function was proposed and the depth image was resolved into several non-intersection sub-level images along z axis. The algorithm complexity was reduced and the postfiltration error was controlled within a given range by filtering and composing sub-level images. The problems of boundary and obstruction were solved by introducing alpha value and composing subimages level by level. This method has been successfully used in computer graphics rendering algorithm and an example of simulating defocus of camera was given.