Abstrato

Research on nondestructive measuring method of tree growth state based on vision technology

Jie Xu, Dawei Qi

This paper aims to improve the accuracy in measurement of tree growth status and to overcome the disadvantages in traditional measurement methods. In this paper, the author starts with parallax processing and combines organically the imaging technologies and vision theories, in order to determine the changes in tree growth status in a period of time on basis of the tree image information acquired at various time points. During the tests, the author calibrates rectangular information points on trees in the image acquired and then acquires tree growth information at such character points by using two cameras, and then carries out comparison studies on the two tree images, in order to calculate and determine the space information of such information points on the trees during a certain time period and to determine the position changes at the information point during such time period. It is found in the tests that for purposes of the calibrated information points, the height and thickness increased respective by 5.63 and 5.75mm by using traditional measurement methods and that the two values are close but do not agree with the actual growth status of trees. However, the heights and thickness increase respectively by 2.4 and 1.5mm by using the visual technology based measurement, of which the height change is 1.6 times larger than thickness change and agrees with actual growth process of trees. Therefore, the measurement methods on basis of visual technology can effectively realize nondestructive measurement of trees, so as to determine the changes of growth status of trees.