Abstract: The cloud typically covers about 50% of the earth surface, affecting the energy balance, water circulation of the earth, and influencing global climate. Cloud cover and cloud base height are the primary elements in the observation of clouds, wherein the cloud amount reflects the regional weather and climate characteristics. However, the operating artificial observation method for cloud amount in China lacks objectivity and costs too much workforce. Therefore, the research and development of an automatic measuring cloud device is extremely necessary.An infrared cloud detector based on the different characteristics of visible and infrared cloud images is developed in order to measure cloud amount. The infrared sensors regularly scan the whole sky under the control of the rotating platform, and stitching all-sky images of infrared brightness temperature. Based on the different characteristics of cloud images shown in infrared, different temperature at different elevation for the cloud point and non-cloud point is considered, combining with real-time infrared brightness temperature and ground environmental parameters. The temperature threshold functions are real-time determined from the zenith to the horizontal, the all-sky cloud distribution and cloud amount are derived using the method of segmentation. This method is effective in reducing the ground environmental parameters and the sunlight on the cloud images, and it can run properly throughout the day. The cloud measuring system can provide real-time analysis of the sky condition, cloud distribution and cloud amount. The system is fully automated and adaptable for all weather conditions, with high temporal resolution and accurate observation time, and it can show the cloud trajectory of the whole period and cloudiness trends.Measurements of cloud system has been performed experimental observation in the synthesize experiment base of Meteorological Observation Center of CMA since May 2011. Through the comparative data analysis with the results of artificial observations and the visible cloud images, this measurement of cloud systems are proved to be able to provide results with high accuracy.