Recently, there has been an increasing need for solar energy production plants due to the expansion of economic activities worldwide and the growing demand for clean and environmentally friendly energy sources to meet these requirements. These fields are usually established in wide areas. Various environmental factors such as dust, snow, pollen, and bird droppings can affect the full penetration of sunlight onto the solar panels, reducing their electricity production efficiency. Therefore, regular inspection and cleaning of solar energy fields are required. However, even regular inspections are usually not sufficient, especially in countries with harsh and unpredictable climates, where energy losses from solar fields are common. In this paper, an automated inspection system based on image processing and deep learning has been designed to ensure continuous monitoring and assessment of the status of solar panels. An Elman neural network has been designed with a focus on improving the image pre-processing algorithm to ensure optimal performance.