Image retrieval systems are essential for efficiently accessing relevant visual content from massive datasets. Over the years, retrieval methods have advanced significantly, transitioning from simple keyword-based systems to content-based models and, more recently, to deep learning-based approaches. This review outlines major categories of image retrieval techniques, including text-based retrieval, content-based image retrieval (CBIR), machine learning-enhanced methods, and current trends in deep learning and hybrid frameworks. The paper also discusses their respective strengths, limitations, and prospects for further research.