| Along with importance of utilization and protection of ecological environment being more and more realized and the rapid development of computer technology, digital photogrammetry technology and virtual reality technology, visualization research of the tree in recent years has become popular in many areas such as forestry, ecology, surveying. The morphology measurement of the tree is particularly important as a key technology in visualization research of the tree. Because of tree’s morphology is complex and variable, there is a big difficulty to determined through a simple traditional measurement methods, the close-range photogrammetry technology for determination of tree morphology provides a theoretical basis and technical support.Experiment was conducted in experimental field of agroforestry system, in Kejing, Zhicheng and Potou township, Jiyuan City, Henan Province, in September 2009 to December 2009, based on Multi-baseline Digital Close-range Photogrammetry System. 5, 6, 7, 8, 9 year-old walnut trees and 5, 6, 8, 9, 10 year-old apricot trees were selected as the research object. The main conclusions are as follows:(1) Compared the data about one 5 year-old apricot tree from tested points by multi-baseline digital close-range photogrammetry and total station, X coordinate of the average absolute error, maximum and minimum values were 0.0016m, 0.0140m and 0.0000m, the average relative error, maximum and minimum values were 0.028%, 0.097% and 0.000%, Y coordinate of the average absolute error, maximum and minimum values were 0.0019m, 0.0160m and 0.0000m, the average relative error, maximum and minimum values were 0.039%, 0.125% and 0.000%, Z coordinate of the average absolute error, maximum and minimum values were 0.0012m, 0.0040m and 0.0000m, the average relative error, maximum and minimum values were 0.016%, 0.048% and 0.000%. According to results of paired samples test, the two methods have no significant difference.(2) We used multi-baseline digital close-range photogrammetry as the independent variable x and the total station measurement value as the dependent variable y, respectively, X, Y, and Z coordinate values of the linear regression fit can be seen that X coordinate value of the slope was 0.9993, intercept was 0.0110, correlation index was 0.9998, Y coordinate value of the slope was 0.9977, intercept was 0.0306, correlation index was 0.9995, Z coordinate value of the slope was 0.9991, intercept was 0.0108, correlation index was 1.0000, the confidence of the coordinate values of the correlation coefficient test reached 99%. The results show that the data of multi-baseline digital close-range photogrammetry is closely related to direct measurement data of the total station, so it can meet the request of the determination of the economic trees in landscaping morphology and the visualization research of the tree.(3) We selected one tree from the various old walnut and apricot trees randomly and obtained the error of the sample trees each axis direction, the error of the horizontal direction and the global observation error, in order to eliminate the artificial error of photogrammetry and data processing, every sample trees were measured three times a self-calibrating bundle block adjustment and average value was gained. The result indicates that average error of x axis direction was 0.0010m, average error of y axis direction was 0.0014m, average error of z axis direction was 0.0004m, average error of horizontal direction was 0.0018m and average global error was 0.0018m, the results show that in all directions errors are relatively small.(4) Taking one 6 year-old walnut tree and one 9 year-old apricot tree as example, main parameters of morphology such as tree height, diameter, height under branch, crown width and branching angle were obtained by using multi-baseline digital close-range photogrammetry system to meet the request of the accurate determination of every growth stage with the tree morphology and the visualization research of the tree. |
PDF Full Text Request