Poisson Image Editing is a new technique permitting to modify the gradient vector field of an image, and then to recover an image with a gradient approaching this modified gradient field. This amounts to solve a Poisson equation, an operation which can be efficiently performed by Fast Fourier Transform (FFT). This paper describes an algorithm applying this technique, with two different variants. The first variant enhances the contrast by increasing the gradient in the dark regions of the image. This method is well adapted to images with back light or strong shadows, and reveals details in the shadows. The second variant of the same Poisson technique enhances all small gradients in the image, thus also sometimes revealing details and texture.
Source CodeThe source code, the code documentation, and the online demo are accessible at the IPOL web page of this article 1 . In this page an implementation is available for download. Compilation and usage instruction are included in the README.txt file of the archive.Keywords: color enhancement, contrast adjustment, Fourier transform
OverviewThe concept of image editing encompasses the operations by which the local content of one or several images is selected and manipulated to create new synthetic images. The simplest such operation is a copy-paste of a part of an image into another. Local contrast or color adjustments after selecting manually a part of the image (for example the shadows) is another variant.Most image editing operations modify an input image I on a rectangular domain R by manually selecting a region Ω in it. The task is to fill in this region by picking information from the rest of the image R \ Ω, or from another image or, in the case of a selective contrast change, from the image in Ω itself. The most recent algorithms to solve this problem are based on partial differential equations. The main challenge in image editing is to avoid suspicious color or texture alterations which would 1 http://dx