Optimization of Geometry Algorithm for DBH of Standing Tree on 2D Laser Detection

Abstract:

It takes an important role to measure the DBH and position of standing tree by a 2D laser detection in current forestry surveying and mapping. A compensation scheme in short distances contraposing the two variables was put forward on three common geometry algorithms (arc length method, tangent method, dual-cosine method), and a compensation angle was used in the scheme. The functionary relation of compensation angles for arc length method and tangent method with distances was researched specifically, the regression analysis was also carried on, and the fitting equations were obtained (R2>0.85 for each fitting equation). Verification test for trunks using the fitting equations in the range of 0~5.5m was carried out in which five poplar trunks were selected. The accuracies of measured values of radius and the distance were all improved by algorithms compensation. Thereinto, the measure precisions of radius values from arc length compensation algorithm and tangent compensation algorithm were respectively increased by 10.6 percent and 10.7 percent, and the mean absolute errors were 4.8mm and 3.8mm; the mean absolute errors of measured values of distance were controlled within 66.0mm and 15.9mm respectively after algorithms compensation. Experiments proved that the two compensation geometric algorithms could both be used as forestry surveying and mapping algorithms in short distances, and the tangent compensation algorithm is better.