The technique of aligning photographs with an artist's drawing is referred to as photo sketch matching. Matching the exact photo from the database is a big challenge, and the best technique must be selected. The focus of the study is to do an analysis of the techniques used in photo-sketch matching. The different techniques analysed in the study include Gabor Shape, Generative Adversarial Networks (GAN), Smart Switching Slime Mold Method, and Lightweight Vision Transformer. The review indicates that no single technique universally outperforms others across all conditions; rather, the choice of method depends heavily on the application context (e.g., forensic sketches, artist-drawn sketches, or automatically generated sketches), suggesting that combining the strengths of classical and deep learning methods, solving domain adaptation, and ensuring scalability for real-world deployment. The novelty of this review lies in the comprehensive synthesis and reorganization of existing research on photo–sketch matching. The work goes beyond by providing a critical comparison of techniques, identifying their strengths, limitations, and applicability, such as in forensics. The results show that in order to prevent overfitting, a large dataset is required for training. The intricacy of the issue requires an extensive set of forensic drawings. The accuracy is affected by multiple factors and can be improved by using hybrid techniques or deep learning approaches.