Devices available on the market to measure particulate matter are too expensive to deploy them with high spatial resolution. In this paper we present a particulate matter measurement device, which is based on the low-cost SPS30 sensor. We show that using a feedforward neural network running on a Raspberry Pi 4, which also considers the ambient temperature and humidity, the measurement results can be calibrated to the ones like the cost-intensive particulate matter measurement device Fidas 200 E for a defined population of particles. We show that after the calibration the measurement results of our low-cost device are comparable to the ones of the cost-intensive device in a comparable particle population. We conclude that the use of feedforward neural networks allows the usage of low-cost particulate matter measurement devices to archive high spatial resolution when monitoring particulate matter concentrations.