Traditional PM2.5 concentration detection equipment generally has problems, such as poor integration of monitoring information and topographic information, single collection of data, and small detection range. Therefore, a novel type of PM2.5 detection system based on terrain classification and UAV remote sensing technology was designed and developed. The system takes the single-chip microcomputer mounted on the UAV as the control core, contains a variety of sensors, data processing and transmission modules and GPS modules, and uses solar panels batteries. Based on this, the system could obtain other relevant data related to PM2.5 concentration data at the same time in actual work (such as wind speed, altitude, air humidity and real-time geographic images around the abnormal point). Then, combined with image recognition technology, the collected data is divided into four categories: industrial area, urban residential area, agricultural area, and water system according to its topographic characteristics. Next, according to the air quality classification standards built into the system, the data is further classified and evaluated. Finally, the processed data is sent to the receiving end of the operator via the wireless transmission module. The novel system has the characteristics of simple operation, wide monitoring range and various data collection, which improves the monitoring efficiency of PM2.5 concentration to a certain extent and has good application prospects.