Applied Physics and Mathematics Annotation << Back AUTOMATED SYSTEM FOR COLOR MARK RECOGNITION IN THE PROBLEM OF OBJECT DETECTION M.V. BOBYR, N.A. MILOSTNAYA, A.A. DORODNYKH The article presents an automated color mark recognition system, which is used to solve problems of detection from the camera lens to the object. A rectangular mark fixed on a rodless cylinder of a mechatronic complex for parts sorting is used as a detection object. The model and a parallel algorithm for determining the area of a recognizable color mark, which is used in the regression model to estimate the distance measurement, are described. Presented to comparative analysis of regression models, performance evaluation, calculations using MAPE and Accuracy indicators. Keywords: automated recognition system, object detection, fuzzy logic, fuzzy clustering, algorithm for determining the area of a recognized mark, regression models, MAPE, Accuracy. DOI: 10.25791/pfim.01.2023.1247 Pp. 03-11.