Bionic vision can be adjusted depending on the scene type. Different scenes and visual aid devices can be suited to increase their effectiveness. For a scene to be translated into an algorithm it first must be categorized. There are some automatic scene categorization models, for instance Chernyak and Stark created one using Bayes’ theorem . Segment features such as aspect ratio and average color are obtained from a test image. Once the scene has been accurately determined by the user the authors want to apply context dependent importance weighting to the image. Image mapping, finding the most important object in an image, is used through several algorithms like lossless compression, military target detection and advertising. The authors performed an experiment where they proposed that scene weighted processing can improve perception of low quality images. Figure 2 was presented to 20 normally sighted people. Results show that scene-dependent importance mapping is a good tool to use in automatic optimization of low quality images. The authors work applies to importance map methods which bring together aspects of image that are known to influence the attention of the human viewer. Such features include shape, size and contrast.