MATLAB-based Gesture Recognition Using HOG Feature Extraction and Euclidean Distance Classification

I conducted gesture recognition experiments using MATLAB. In this implementation, I employed the Histogram of Oriented Gradients (HOG) method for robust feature extraction and used Euclidean distance as the classification metric to determine gesture matches. Gesture recognition technology analyzes human hand movements to interpret and understand human intentions. The MATLAB implementation involved key functions such as extractHOGFeatures() for capturing gradient orientation distributions from preprocessed hand images. The feature vectors were then compared using pdist2() function with Euclidean distance calculation to measure similarity between test gestures and reference templates. This approach enabled accurate gesture identification by finding minimal distance matches in the feature space. This research methodology provides a solid foundation for developing more intelligent and interactive human-computer interfaces. The code structure allows for easy integration of additional gesture templates and can be extended to real-time recognition systems using webcam input and preprocessing pipelines for hand segmentation and normalization.