An Object Recognition Procedure Based on the Image with the Specified Color Features in the Aquatic Environment

The article considers an approach to recognize an underwater object from its image obtained with the help of a technical vision system. The color of the object is used as a feature to distinguish the object in the image. The detection problem is solved for an AUV, which follows a trajectory within a monitoring mission to detect and classify an object with known characteristics. Based on the proposed method, a program has been developed to provide the AUV with an image of the detected object and its characteristics. The method is based on the analysis after transcoding from RGB to HSV. According to the experimental data, the brightness interval of the image, which is considered to be not corrupted and in which it is possible to detect objects, is determined.