On the removal of shadows from images

Abstract

Abstract-This paper is concerned with the derivation of a progression of shadow-free image representations. First we show that adopting certain assumptions about lights and cameras leads to a 1-d, grey-scale image representation which is illuminant invariant at each image pixel. We show that as a consequence, images represented in this form are shadow-free. We then extend this 1-d representation to an equivalent 2-d, chromaticity representation. We show that in this 2-d representation, it is possible to re-light all the image pixels in the same way, effectively deriving a 2-d image representation which is additionally shadow-free. Finally, we show how to recover a 3-d, full colour shadow-free image representation by first (with the help of the 2-d representation) identifying shadow edges. We then remove shadow edges from the edge-map of the original image by edge in-painting, and we propose a method to re-integrate this thresholded edge map, thus deriving the sought-after 3-d shadow-free image.

Dealing with Multi-Scale Depth Changes and Motion in Depth Edge Detection

“…In this case, detection of depth edges can be achieved by using algorithms that detect shadows [6] or separate illumination from reflectance using a single image [17].…”

“…Finlayson et al [6] proposed a method to remove shadows from images by deriving a 1D illumination invariant image representation based on log-chromaticity coordinates. Tappen et al [17] use color information and a classi- fier trained to recognize gray-scale patterns in order to classify image derivatives as being caused by reflectance or illumination changes.…”

“…Research on shadow segmentation and intrinsic image computation from a single frame [17,6] is very important to achieve a more general solution to the problem. We also believe that solutions involving new camera setups would be possible.…”

Removing Shadows from Video

“…It should be noted that our method requires the detection of only one shadow with the highest confidence. Once this is achieved, we can use it as a Reference Shadow to estimate the parameters necessary to generate a grey-scale invariant image [1].…”

“…To solve this issue, various studies use Gaussian to model shadows dynamically; However, good results are achieved only when the scene meets a series of assumptions [21]- [23]. Another important study on this topic is published by Finlayson et al [1], who introduced the concept of greyscale invariant image and presented a computational model to estimate the invariant image. Same authors published another study and created invariant image by Entropy Minimization [24].…”