Researching a learning algorithms efficiency for colorization by using a multilayer perceptron

The article describes the software for grayscale images colorization by using a multilayer perceptron. The system has been developed by RAD (Rapid Application Development) methodology. The colorization is performed by multilayer perceptron learned by the back-propagation algorithm, the back-propagation algorithm with a moment, the back-propagation algorithm with a moment and dynamic learning coefficient. The article presents the results of the research representing the quality of image colorization.