Recent advances in wildlife monitoring have created new opportunities for non-invasive species identification. Our research investigates an innovative approach to gender classification in blackbucks through pugmark analysis, with particular emphasis on developing efficient methods for identifying regions of interest in the blackbuck pugmark (BP) dataset. We conducted a comprehensive evaluation of three leading machine learning algorithms—logistic regression, support vector machine (SVM), and random forest classifiers—to analyze our extensive pugmark image repository. Through rigorous testing, the random forest classifier emerged as the superior performer, demonstrating remarkable accuracy across multiple evaluation metrics. The significance of our findings extends beyond establishing a reliable baseline for the BP dataset. Our results offer crucial insights that could revolutionize automated footprint recognition systems in wildlife studies. Perhaps most importantly, the methodological framework we’ve developed shows considerable promise for adaptation across different species, potentially transforming how we approach wildlife monitoring and conservation efforts.