Investigating Sensor-based Condensation Detection in Network Cameras using Psychrometry

The presence of fog, or condensation, poses significant challenges in various fields, including surveillance cameras. The main issue is that fog obstructs the cameras view, essentially rendering the camera ineffective for surveillance. Currently, preventative measures such as heating the camera is a remedy at the cost of significantly increased power consumption. This master’s thesis is aimed to enable smarter fog prevention by detecting when fog is present. This is achieved by exploring and testing several fog detection techniques from different industries to develop a prototype capable of detecting the presence of fog. Four techniques: hygrometry, capacitance, resistance and infrared were initially investigated where hygrometry was decided as the most promising approach. This involves measuring temperature and humidity inside the camera in order to estimate the dew point temperature which is compared to the surface temperature where fog is prevalent. The temperature comparison represents a calculated indication of whether fog formation is theoretically possible. The results were encouraging, indicating promise that the concept could effectively detect fog. The result and conclusion of this master’s thesis is a developed proof of concept which is able to detect the presence of fogging in outdoor network cameras. The solution consumes an insignificant amount of energy compared to the potential energy savings from detecting fog, demonstrates a potential to be reliable and finally, holds significant promise as an effective detection method.

The work has been carried out at Axis Communicatios AB, Lund .

Supervisors: Albin Berggren, Axis, Gunnar Lindstedt, IEA

Examiner:  Johan Björnstedt, IEA