Application of La*b* Color System for Evaluation Images Quality of Rye-Wheat Bread

The quality of input images is essential for all processes related to extraction of information about captured objects and their characteristics. Thus, current research focuses on evaluation the quality of images of rye-wheat bread regarding the influence of lighting conditions. For the purpose of the experimental work two types of rye-wheat bread that are developed to have a positive effect on glucose metabolism, are used. The examined bread samples are captured using two types of lighting- one based on artificial LED light and second which is based on “like a daylight” conditions. Colors of all images are analyzed using confidence intervals to evaluate their statistical parameters. As additional tool for image analysis the kNN classifier is used in order to assess how lighting conditions influences on recognition of captured type of bread based on its images. The research findings can serve improvement the methods of computer vision regarding selection the lighting conditions.