In order to improve the efficiency of power transmission line inspection by Unmanned Aerial Vehicle (UAV), a new method was proposed for detecting broken transmission lines and defects of foreign body based on the perception of line structure. The transmission line image acquired by UAV was easily influenced by the background texture and light, the gradient operators of horizontal and vertical direction which can be used to detect the line width were used to extract line objects in the inspection image. The study on calculation of gestalt perception of similarity, continuity and colinearity connected the intermittent wires into continuous wires. Then the parallel wire groups were further determined through the calculation of parallel relationship between wires. In order to reduce the detection error rate, spacers and stockbridge dampers of wires were recognized based on a local contour feature. Finally, the width change and gray similarity of segmented conductor wire were calculated to detect the broken part of wire and foreign object defect. The experimental results show that the proposed method can detect broken wire strand and foreign object defect efficiently under complicated backgrounds from the transmission line of UAV images.
For the image processing, computer vision and 3D rendering have the feature of massive parallel processing, the programmability and the flexible mode of parallel processing on the Polymorphic Array Architecture for Graphics (PAAG) platform were utilized adequately, the parallelism design method by combing the operation level parallelism with data level parallelism was used to implement the OpenVX Kernel functions and 3D rendering pipelines. The experimental results indicate that in the parallel implementation of image processing of OpenVX Kernel functions and graphics rendering, using Multiple Instruction Multiple Data (MIMD) of PAAG in parallel processing can obtain a linear speedup that the slope equals to 1, which achieves higher efficiency than the slope as nonlinear speedup that less than 1 of Single Instruction Multiple Data (SIMD) in traditional parallel processing of the Graphics Processing Unit (GPU).
To solve the problem of detecting highway-vehicle abnormal behavior such as retrograde motion, parking and abnormal trajectory, this paper presented a bottom-up detection algorithm based on lane model. First, the lane line and vanishing point were found out by line's continuity and collinearity, and the lane model was automatically established. Second, a region-overlap graph was established by motion prediction and KLT feature tracking to indicate region relationship of the object in the detecting and tracking process. In this graph, the corresponding relationship and reliable trajectory was made by merging or splitting the target region. The target region was ruled by posterior relationship. At last, the vehicle's location was transformed based on vanishing point. The trend of target motion was judged and its location or velocity was calculated in the lane model with sliding window to decide vehicle's behavior. The experimental results show that the proposed algorithm has more than 80% detection rate for car incident in different weather or traffic environment. This algorithm is capable to detect vehicle abnormal behavior on highway for real-time application.