Monday, 23 January 2012
An Inexpensive Automated Webcam-Based Laser Snow Depth Sensor Design
Hall E (New Orleans Convention Center )
The depth of snow on the ground is a critical environmental variable that impacts local water resources, agriculture, ecology, transportation, and recreation. Automated snow depth sensors are operational worldwide as part of long-term meteorological stations, scientific studies, and environmental monitoring. At present, ultrasonic sensor designs are the industry standard largely because of the system's low power requirements. However, these sensors must account for dependence of the speed of sound through air on the ambient air temperature, which introduces a potential error source. Additionally, ultrasonic depth sensors are particularly insensitive to measuring slight snow accumulation on bare ground; events that critically change the surface conditions (e.g. temperature, albedo, roughness, traction). A sensor system that relies upon optics rather than acoustics offers more flexibility in the type of data it could provide and increased sensitivity to slight changes in surface characteristics. Optical sensors have been developed previously with great success, but these instruments tend to be complex and expensive, thus limiting their practicality for widespread use. We present preliminary results from a prototype webcam-based laser snow depth sensor. The design uses simple geometry between a downward-looking camera and laser diode aligned in parallel and at a known spacing. The distance between the sensor and the ranged surface is inversely related to the distance, in pixels of the digital photograph, between the projected laser point and image center. The sensor exhibits a measurement error of <1% to a range of >4 meters. Both visible and near-infrared designs have been tested. In addition to snow depth, the infrared design is capable of logging the fraction of snow cover in the field of view under any ambient light conditions. The sensor is designed to be robust, miniature, precise, wireless, network compatible, and with internal memory and a rechargeable battery. The units would be capable of remote, network deployments of multiple sensor nodes without ties to expensive power and data logger systems. The design shows promise as a new, inexpensive method of snow depth data collection.
Supplementary URL: