A Method for Determining Obstacle Contours Using Wearable Video Cameras for Orienting Blind and Visually Impaired People in Urban Environment

Abstract

The essence of the method is based on the division of the video camera matrix into submatrices, in which the Euclidean distance is determined between adjacent rows and columns. If there is a gradient of pixel intensities exceeding a certain threshold, this leads to a jump in the Euclidean distance, which indicates the presence of an element of the contour of the entire object in a separate submatrix. The quality of such a contour determination depends on the sizes of the submatrices and the selected threshold. To expand the capabilities of the method in estimating the complex shape of objects, a combined submatrix method is proposed, in which contours are determined at different submatrix sizes and threshold values, and then superimposed on each other. The sizes of two different submatrices are determined based on previous research. The possibilities of optimizing such sizes using a genetic algorithm are considered. The method provides the highest similarity of contours to the original image of steps compared to other methods, and the accuracy of determining the contours of a rectangle, triangle, and circle is comparable to the accuracy of such methods.



Received: 14 December 2025 | Revised: 5 March 2026 | Accepted: 17 March 2026



Conflicts of Interest

The authors declare that they have no conflicts of interest to this work.



Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.



Author Contribution Statement

Oleksandr Poliarus: Conceptualization, Methodology, Validation, Investigation, Resources, Data curation, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration. Yevhen Poliakov: Software, Validation, Formal analysis, Resources, Data curation, Writing – original draft, Visualization.