This paper introduces a black-and-white image colorization model based on deep learning techniques and OpenCV by combining available open-source toolkits. The model uses the Lab color space, in which a single grayscale image (the luminance or L channel) that is processed by a convolutional neural network (CNN) predicts chromatic values (a and b channels). The colorized version of the image is, therefore, generated by merging it with the original L channel. This approach makes use of deep learning in order to enhance the quality and performance of the reconstructed images. Experimental results verify that the proposed method is both valid and flexible and, hence, can vividly restore color from monochrome photographs in such domains as historical photo restoration and artistic creation.