Solar energy, because of its unique advantages like infinity and low pollution, has been considered as a major renewable source of energy over the last two decades and is widely being used in urban, industrial, and agricultural sectors. The performance of solar panels in energy production is highly depends on their irradiation gain. The higher the panels’ angle perpendicular to the solar irradiation, the more increase in the panels’ irradiation gain. Solar trackers are the most appropriate technology make it possible to automatically adjust the solar panels in the right angle respect to the solar irradiation. With respect to the type of control unit, solar trackers are classified into two major categories: passive and active. Passive systems work usually based on thermal expansion of a gas fluid with low boiling point like Freon or on shape memory alloys. These systems do not use any electrical circuit, but relatively are less efficient and do not operate at low temperatures. In contrast, active systems use electronic circuits and are more accurate than the passive systems. Sensing elements in the active systems are usually anti-parallel connected photo-resistors and PV solar cells. While launching the solar trackers with these kind of sensors is without any complexity, as a big disadvantage, these sensors lose their performance when the sky is cloudy. To overcome this challenge, scientists in the recent years have focused on developing machine vision-based approaches to detect the position of the sun in the sky even if the sky is cloudy. A vision-based solar tracker usually consists of a camera, electronic circuits with microcontroller, and stepper motors. The position of solar in the sky is determined by analysis of images acquired from the sky. Images are first subject to some pre-processing treats such as segmentation and binarization. Then, after camera calibration, the position of sun is calculated using the morphological operator and regression analysis.
