Intensity edge processing of Moroccan sign language video frames using Sobel edge detection and thresholding techniques

This paper presents an edge intensity processing framework applied to Moroccan sign language video frames using Sobel edge detection and hysteresis thresholding. The methodology consists of pre-processing video images by converting them to grayscale, then applying the Sobel operator to extract edge features. Intensity thresholds are set to classify pixels as low, high or hysteresis-based edges, to improve gesture boundary detection. The framework calculates average edge intensities per row and visualizes them for comparative analysis. Experimental results demonstrate the robustness of the proposed approach in identifying significant gesture features, supporting potential applications in sign language recognition and translation systems. The study highlights the effectiveness of contour-based methods for gesture based pattern selection in real-world sign language datasets.