An Efficiency-Driven Deterministic Optimization Approach for Sensor Placement in Image-Based Forest Field Measurement
In forest management, one key objective of forest field measurement is to measure the Diameter at Breast Height (DBH) of each tree in a specified area. Nowadays the widely employed way is to measure the trees manually one by one which usually takes several days. However much work can be considerably saved by adopting image based measurement approach which includes several steps. Among these steps, careful planning of the sensor placement is an essential preparation step which has significant impact on the effectiveness of the following steps. In this paper, a concept named Trees Per Location (TPL) is proposed to evaluate the efficiency in sensor placement. Based on TPL, we present a novel automatic sensor placement algorithm suitable for image based forest field measurement. The key feature of the proposed algorithm is that the impact of the camera orientation on optical constraints is attenuated due to the fact that in outdoors, the orientation of camera is not easy to control compared with the location of camera. Our method generates a plan composed of a series of sensor viewpoints and a shortest path that traverses each viewpoint exactly once. The plan guarantees that the total number of images needed to be taken is minimum and the travel distance of the path is the shortest while our plan satisfies the constraint that each tree appears in at least one image without being blocked by any other trees. Experiments are carried out on a sample forest from the PlotNet database and a real forest in order to compare the proposed TPL with other mainstream algorithms and validate the proposed method for DBH measurement.
KeywordsSensor Placement Kinect Sensor Camera Orientation Sample Forest Optical Constraint
Unable to display preview. Download preview PDF.
- 1.PlotNet Forest Database, http://eco1.ees.hokudai.ac.jp/plotnet/home
- 2.Banta, J., Abidi, M.: Autonomous placement of a range sensor for acquisition of optimal 3-d models. In: Proc. of Industrial Electronics, Control, and Instrumentation, vol. 3, pp. 1583–1588 (1996)Google Scholar
- 6.Khan, F., Khan, N., Inayatullah, S., Nizami, S.: Solving TSP problem by using genetic algorithm. IJBAS 9(10), 79–88 (2009)Google Scholar
- 7.Li, Y., Liu, Z.: Uncertainty-driven viewpoint planning for 3d object measurements. In: Proceedings of the IEEE International Conference on Robotics and Automation, ICRA 2003, vol. 1, pp. 127–132 (2003)Google Scholar
- 8.Reiterer, A., Lehmann, M., Fabiankowitsch, J., Kahmen, H.: Quality control for building industry by means of a new optical 3d measurement and analysis system. In: Observing our Changing Earth. International Association of Geodesy Symposia, vol. 133, pp. 771–779. Springer, Heidelberg (2008)CrossRefGoogle Scholar