Air quality sensors are an emerging new technology of the air quality monitoring. Their main advantage is that they are significantly cheaper monitoring device compared to standard monitoring equipment. Cheap, mass produced sensors have a potential to form much more dense monitoring networks and provide more detailed information about air the pollution distribution. The drawback of the sensor air pollution monitoring is the lower quality of measurements than standard monitoring equipment. Air pollution sensors measurements quality is known to be negatively influenced by meteorological factors, such as temperature or humidity. Neural networks are potentially valuable technique of the monitoring data processing, to transform sensor measurements complemented with meteorological data to more accurate estimations of pollutant concentrations. Co-located measurements at the three monitoring sites in the Ostrava region in Czechia proved that although the PM10 and PM2.5 measurements are relatively highly correlated (0.8-0.9) the measurements need to be adjusted because sensor measurements, based on location, under- or overestimate the particulate pollution. Neural network proved to be the method which can significantly raise the quality of measurements. The neural network postprocessing significantly raised precision of measurements as well as the correlation between professional monitoring station and sensor measurements (approx. to 0.95) which make cheap sensor data reaching sufficient quality for air quality monitoring.