Architectural and technical drawings are prevalent in libraries, specialized architectural collections, and the archives of public or construction companies. The copies of original drawings of the early 20th century are either photographic or photomechanical reproductions and have distinct requirements for long-term storage and stability. The identification and classification of these materials and their associated processes are crucial for developing effective preventive conservation strategies. This study aims to establish an efficient methodology that leverages the advantages of image processing and computational analysis for the recognition and classification of architectural photoreproductions. Traditional identification methods based on stereomicroscopic observation are re-evaluated using machine learning, focusing on characteristics such as color and surface texture of printed media and substrates. Digital microscopy is employed to examine original prints from a technical company's archive in Greece. The resulting digital images are analyzed using computational pattern recognition techniques that emphasize the unique features of each print type. The findings are cross-referenced, and the resulting variables contribute to the creation of an effective and accurate identification system for both photographic and photomechanical prints.