Optical solution for particulate distribution estimation

This paper describes a simple sensor which is designed to sample the air and extract the distribution of particulate it contains. The proposed system employs a normal fiberglass filter, a small video camera and a RaspberryPI to process the image and extract the particulate distribution. The fiberglass filter is capable of stopping particulate down to sizes of 2.5 μm, i.e. what is usually refereed to as PM2.5 particulate and employs a small pump to increase the sensitivity. A motor is used to change the filter interested by the particulate and to restart the measurement. Depending on the amount of particulate in the air, the sampling can be restarted every day or more frequently. The proposed system has been characterized by observing the filter using a field-emission electronic microscope showing that the measured distribution is correct. The overall cost of the proposed system is of less than 100$.