This work is part of an effort for the development of a Knowledge-Vision Integration Platform for Hazard Control (KVIP-HC) in industrial workplaces, adaptable to a wide range of industrial environments. This paper focuses on hazards resulted from the non-use of personal protective equipment (PPE), and examines a few supervised learning techniques to compose the proposed system for the purpose of recognition of three protective equipment: hard hat, gloves and boots. In the KVIP-HC, classifiers, feature images and any context information are represented explicitly using the Set of Experience Knowledge Structure (SOEKS), grouped and stored as Decisional DNA (DDNA). The collected knowledge is used for reasoning and to reinforce the system from time to time, customizing the service according to each scenario and application. Therefore, in choosing the classification methodology that best suits the application, processing time for training (once the system will be eventually reinforced in real time), accuracy, detection time and the predictor sizes (for the purpose of storing data) are analyzed to propose the most reasonable candidates to compose the platform.