Tracking and modeling raindrops in video sequences for assessing precipitation

Disdrometer system is a kind of rainfall measurement tool usually developed to evaluate the raindrop size distribution and velocity of falling hydrometeors. To improve the disdrometer systems, a newly image-based technique for analyzing the raindrops was proposed in the study. This system mainly composed of three components: an optical unit with a light source; an image acquisition unit with a camera and a blackboard to grab the images; the last part is a data processing unit for estimating the collected data. To test the system, we have used glass balls and a liquid dropper to generate free-fall drops in the experiment. Subsequently, the high-speed camera captured images of the falling drops in 90 fps. Especially, the exposure time was set at 2000 ms, which resulted in the falling drops with small blurring in the assessment. We concluded that the proposed image-based disdrometer system can provide an efficient approach for the precipitation measurement by analyzing raindrop features. Moreover, this apparatus have measured micro-physical properties for quantifying the rainfall status as it only requires basic maintenance for long-lasting observation in meteorology especially in large scale areas and distributed scopes.