Kalyta, Oleg2022-04-192022-04-192022-04-14Kalyta O. Information technology of facial emotion recognition for visual safety surveillance. Computer systems and information technologies. 2022. Vol. 4, No 1. Pp. 54-61. DOI: https://doi.org/10.31891/csit-2022-1-7https://elar.khmnu.edu.ua/handle/123456789/11888Over the past decades, information safety systems for visual surveillance based on emotion recognition have become vital. In this work, we propose information technology that serves as a means of recognizing emotions for visual surveillance systems that meet safety requirements, particularly for low-resolution video surveillance cameras. The use of information technology means the consistent application of the method of geometric interpretation to extract facial features and the method of hyperplane classification to identify changes in emotional states. The input data of the proposed information technology is a set of videos with detected faces that reproduce primary emotional states on them. Based on the calculated quantitative characteristics of emotional states, emotional states by facial expressions were identified. The proposed information technology’s efficiency and practical significance are proved by its comparison with analogs according to statistical estimates. As such, the proposed approach achieved the classification accuracy of 91% and surpassed the analogs in the F1-norm, scoring 76%. Furthermore, utilizing straightforward mathematical operations in facial geometric feature representation and hyperplane classification within the information technology considerably reduced the computational cost over the analogs.eninformation technologyemotion recognitionfacial feature extractiongeometric featuresvisual safety surveillancehyperplane classificationСтаття004.932+004.023