We develop a multi-angular imaging power line inspection system. Its main objective is to monitor the relative distance
between high voltage power line and around objects, and alert if the warning threshold is exceeded. Our multi-angular
imaging power line inspection system generates DSM of the power line passage, which comprises ground surface and
ground objects, for example trees and houses, etc. For the purpose of revealing the dangerous regions, where ground
objects are too close to the power line, 3D power line information should be extracted at the same time. In order to
improve the automation level of extraction, reduce labour costs and human errors, an automatic 3D power line
reconstruction method is proposed and implemented. It can be achieved by using epipolar constraint and prior
knowledge of pole tower's height. After that, the proper 3D power line information can be obtained by space intersection
using found homologous projections. The flight experiment result shows that the proposed method can successfully
reconstruct 3D power line, and the measurement accuracy of the relative distance satisfies the user requirement of 0.5m.
In this paper, a mathematic model for POS based bundle adjustment is introduced. The model is made up of four types of linearized observation equations. The intention of the POS based bundle adjustment is to minimizing the error between the four types of observed value and its model value. We use the Levenberg-Marquardt algorithm to achieve this purpose. Our work is supported by China 863 program titled 'airborne multiangular imaging technique in power line inspection' (AMPLI). The purpose of this program is to monitor the relative distance between the power lines and the objects beneath them with accuracy as high as 0.5 meters. A number of high-resolution images must be captured along the power lines to ensure the accuracy. Based on an automatic matching method proposed by other team members in this program, hundreds of homonymous points can be extracted in one image. About 30 to 50 images are used in one block adjustment. As a result, large number of unknowns will contribute to the minimized error, and numerous equations should be solved. So, the minimization algorithm must incur the high computational costs in the problem. Fortunately, the normal equations reconstructed from the observation equations above exhibiting a sparse block structure. Considering the sparse characteristic of the normal equation, we propose a sparse bundle adjustment method based on Levenberg-Marquardt algorithm to save computation cost. A software package is developed based on this algorithm. A comprehension test was performed to investigate the performance of the algorithm. We used a data set provided by a field experiment in Wuhan, China. It is found that our algorithm showed both high accuracy and high efficiency in the test.