In today's world, where environmental problems are becoming more and more urgent, the search for alternative energy sources is becoming a priority. One of the most promising areas is solar energy. Solar energy is a renewable energy source, which makes it attractive for use in various areas, including power generation, heating and cooling of buildings, and transport. The development of solar energy can contribute to solving a number of environmental problems such as pollution and climate change. However, solar panel equipment is subject to various types of defects and contamination. Defects can adversely affect the performance and efficiency of solar panels, so their detection is critical to improve the reliability and durability of photovoltaic power generation systems. Effective fault finding can minimise energy losses, improve system reliability and equipment life, and reduce maintenance costs. In addition, improved performance of electrical equipment contributes to the sustainable development of alternative energy, thus reducing dependence on conventional energy sources and reducing greenhouse gas emissions. The paper presents an overview of existing methods for detecting various solar panel faults using computer vision and deep learning techniques. Infrared thermography (IR), electroluminescence (EL) imaging, or visible spectrum imaging can be used to find the faults. This paper includes an analysis of the advantages and disadvantages of existing methods for finding defects and contamination in solar panels, discusses the factors affecting their performance, and presents conclusions for possible future research in this area.
