Spiking Neural Networks (SNNs) stand out as the third generation of artificial neural networks that mimic the event-driven and sparse computation principles of biological nervous systems. The high energy consumption and computational requirements of traditional deep learning models make SNNs an attractive alternative, especially in real-time and resource-constrained applications. This study comprehensively examines the applications of SNNs in the field of image processing in light of fundamental concepts and current literature. The study first describes neuron models and information encoding strategies; then, SNN-based approaches in various image processing tasks such as edge detection, image enhancement, object detection, and classification are detailed. In addition, SNN training methods, datasets used, and neuromorphic hardware platforms are discussed. Finally, current challenges such as the differentiability problem, the lack of benchmarks, and the need for algorithm-hardware co-design are discussed; and the development of systems integrated with continuous learning and neuromorphic sensors is presented as future research directions.