Warning: Low Voltage Power Line > 자유게시판

• 목록
• 답변
• 글쓰기
• 게시판 리스트 옵션
  • 수정
  • 삭제

The Gabor-YOLO algorithm in this research is composed of an adaptive foreground extraction module based on the Gabor operator, an improved YOLO network based on attention mechanism, and a reasoning module based on contextual info. Aiming at the lengthy and slim linear construction of power strains in the image, a double feature layer extraction structure and a pyramid multi-scale fusion module are adopted. The characteristic map output by the primary residual module is used as a low-order geometric function map, and the feature map output by the third residual module is used as a high-order semantic function map. The community combines the extraction of goal characteristic information with the realization of target classification, and end-to-end coaching is achieved for the first time. However, the sting detection operator has the advantages of simplicity and efficiency, so it can be used to detect the sting of the picture first and establish the candidate pool of lines pixel points, which can't solely enhance the accuracy of power strains extraction but additionally significantly scale back the computation of the subsequent convolutional neural community. At the identical time, for bettering the adaptability of lightweight backbone to low-voltage overhead traces, on the one hand, a convolutional block attention module was added that combines channel attention and spatial attention, and the feature fusion department was adjusted to determine.

As shown in Figure 5, compared with excessive-voltage transmission traces, overhead distribution lines of low-voltage distribution network journey in streets and residential buildings, with very serious background interference and shielding. Transmission strains should not have any coloration code in their plastic masking. So as to confirm the advantages and effectiveness of the algorithm in our study, precise aerial pictures of an actual low-voltage overhead traces in Wuhan, Hubei, China were used to conduct the simulation verification experiment, with the following software program and hardware platforms: Intel Core i5-10400F@2.90GHz×; 6 CPUs, NVIDIA GeForce RTX 2060, Ubuntu 16.04LTS working system, Pytorch deep studying framework, and Nvidia Jetson Xavier NX clever edge machine. The R-CNN series of algorithms are within the order of their development. Figure 8. Results of varied algorithms. Within the inference module, K-means clustering was carried out on the coordinates of all auxiliary targets, and after the power distribution channel was obtained, and the IOU calculation was performed with the ability line, to obtain the ultimate power strains extraction results. Figure 7. Results of every step in our mannequin. The function fusion mannequin in this examine is different from the conventional model.

Combined with the truth that the power strains itself involves much less high-order semantic features and the velocity requirement of power strains extraction, the lightweight spine was adopted within the examine. These are typical small targets with very weak features of strains and insulators. On this basis, the improved YOLO community on this part can determine energy strains and auxiliary objects, for instance, insulators and C-clamps. If there is only one single clustering result, IOU calculation could be carried out based on the tower area and power traces, to finish the computerized energy traces extraction of the algorithm. Therefore, these unique gadgets are detected together and used to assist extraction of energy traces. Therefore, we made a new development and efficiently applied Gabor-YOLONet to low-energy terminals Nvidia Jetson Xavier NX and Rockchip RK3399pro. The important thing structure of the residual convolutional neural community is a skipping connection structure. F (x) represents the ahead branch, and x represents the skipped connection department. Grayrepresents the grey value after gray processing; Rrepresents the gray worth of the red element; G represents the gray value of inexperienced element; and Brepresents the gray value of the blue element. As could be seen from Figure (b), the mAP value of the algorithm in this research has risen to 0.6 (point A in the figure) in the primary 50 rounds, and finally stabilized at round 0.9 (level B in the determine), reaching excessive recognition accuracy.

It can be seen from Figure (a) that the initial loss worth of the mannequin on this research was 29.Ninety three (level A within the figure), decreased to about 6.10 (level B in the figure) in the primary 50 rounds of iteration, and the loss value finally decreased to about 4.13 (point C within the determine). Initially, they will never set up power strains in trees or some other place where they'll come into contact with the rest. The RCNN first scans the enter picture with the selective search algorithm to extract candidate boxes, then scales all candidate packing containers to a fixed pixel size by means of normalization, after which inputs them into the convolutional neural network to unify the length of the feature vector. However, Faster RCNN (Ren et al., 2015) makes use of the regional candidate community instead of the selective search method and obtains the adjusted candidate box by setting anchor containers of different scales and combining with the convolutional neural community. The standard is then extracted, the characteristic vector in every candidate region of the SVM binary classifier is supplied in keeping with the goal class to categorise the corresponding number, and the goal place info is obtained through regression to complete the goal detection.

If you have any sort of concerns regarding where and just how to make use of low voltage power line, you can call us at our own site.