Download (PDF 825 KB): https://research.fs.usda.gov/download/treesearch/7659.pdf
Abstract
A multisensor video system has been developed incorporating a CCD video camera, a 3-axis magnetometer, and a laser-rangefinding device, for the purpose of measuring individual tree stems. While preliminary results show promise, some changes are needed to improve the accuracy and efficiency of the system. Image matching is needed to improve the accuracy of length measurements spanning multiple video frames. Formulation of a knowledge base can enhance the speed and robustness of the matching algorithm. This paper will demonstrate how characteristics of sub-canopy video collection and parameters of the video system can be used to formulate a suitable image matching algorithm.
Citation
Clark, Neil A. 2001. Formulating an image matching strategy for terrestrial stem data collection using a multisensor video system. Proceedings, 18th Biennial Workshop on Color Photography and Videography in Resource Assessment. 9 pp.
