Objectives The paper proposes a technique that allows the main types of clouds and cloud cover parameters (ceiling, direction and speed of cloud movement) to be determined at minimum cost via observations from the Earth’s surface. The aim is to avoid using meteorological information derived from official sources, which may not always be available for a specific place and time of observation.Method The method is based on the well-known Ippolitov formula that connects air humidity on the Earth’s surface with the cloud ceiling, whose empirical coefficients depend on the type of the cloud cover. In order to determine this type, a method based on elliptic cloud contour approximation is proposed.Results Additional classification features were the number of approximating ellipses in the image frame, the proximity of their mutual location, the total area that they occupy in the frame and the area occupied by the largest ellipse. The speed and direction of the clouds are determined through shifting the common key points in image frames obtained over a small fixed time interval. Each point is described by a descriptor calculated using the SURF method from the open image processing library OpenCV.Conclusion Despite the simplicity of the required software and hardware (web camera, humidity sensor, and OEM – Arduino module), this method provides a definition of these parameters with an error within 20%. Monitoring systems based on this method may be used in private automatic weather stations, such as the Wunderground, for obtaining more information that could improve the accuracy of weather forecasts, as well as for research in meteorology and climatology.