Transmission Line Inspection Image Recognition Technology Based on YOLOv2 Network

With the planning and implementation of national smart grid construction, on-line monitoring technology, as one of the key technologies of smart transmission, is also developing towards intelligence. With the construction of smart grid, the number of UHV long-distance and large-capacity transmission lines is increasing, and the load density of the grid is also increasing. As an important component of power system operation, the safety and reliability of power lines are very important. At the same time, with the popularization of helicopter inspection of power transmission lines, it not only provides opportunities but also challenges for automatic detection of defects in power transmission lines by image processing and recognition technology. How to effectively obtain basic information of power line operation from power line images obtained by helicopter inspection has become a research hotspot. Using automatic extraction method of image recognition can not only improve work efficiency and detection accuracy, reduce patrol personnel, but also overcome various defects of manual methods, and finally improve the safety of power line patrol and the accuracy of detection results.

Through the recognition of ocean sediment sonar images, the texture in the image can be classified, which provides an important basis for the classification of ocean sediment. Aiming at the problems of low efficiency, waste of human resources, and low accuracy in the traditional manual side‐scan sonar image discrimination, this paper studies the application of image recognition technology in sonar image substrate texture discrimination, which is popular in many fields. At the same time, considering the scale complexity, diversity, multisources, and small sample characteristics of the marine sediment sonar image texture, the transfer learning is introduced into the image recognition, and the K‐means clustering algorithm is used to reset the prior frame parameters to improve the speed and accuracy of image recognition. Through the experimental comparison between the original model and the new model based on transfer learning, the AP (average precision) value of the yolov3 model based on transfer learning can reach 84.39%, which is 0.97% higher than that of the original model, with considerable accuracy and room for improvement; it takes less than 0.2 seconds. This shows the applicability and development of image recognition technology in texture discrimination of bottom sonar images.

With the rapid development of technology, artificial intelligence is gradually applied in the field of image processing. Traditional image processing and recognition takes human extraction of features as a means to describe and characterize low-level edge information in images, but a large number of features extraction causes certain troubles to image recognition, and it is difficult to extract deep-level information. In recent years, deep learning, as an important emerging technology in the field of image recognition, has broad application prospects, and the research on image recognition technology has important theoretical value and practical significance to promote the development of computer vision and artificial intelligence. It opens up a whole new direction in image processing research, thus meeting the demand for efficient image processing. This paper reviews the applications of deep learning in image recognition, analyzes the classical convolutional neural network model, and summarizes the applications of deep learning in the fields of license plate recognition, face recognition, and medical image recognition.

Transmission line is the key component of the power system, and its working condition is very important for the safe operational aspect of the power system. Transmission line inspection plays an essential role in ensuring the power system can be operated safely, reliably, effectively. With the escalation of the scale of the national power system, the demand for inspection of power transmission lines is gradually increasing. Due to the shortcomings of traditional human inspection, such as time-consuming, laborious, and not strong real-time, UAV electric inspection develops rapidly. As an important index for the security and reliability operation of transmission lines, the size of sag directly affects the security and stableness of transmission lines, so it’s an essential part for UAV inspection to measure sag. This research studied the sag measurement of transmission lines by relying on the UAV platform, with the baseline of the requirements of transmission line inspection. On the basis of image preprocessing, this paper adopts the method based on 3D reconstruction, introduces the camera calibration method to obtain parameters, calculates the 3D coordinates of the highest point of the tower and the lowest point of the line in the image from the pixel coordinates, and finally calculates the sag value according to the concept of the sag. Compared with traditional measurement methods, this study has smaller error, less calculation and more convenient measurement.

Background: With the rapid development of digital twinning technology, the compatibility of digital twinning technology to other technologies is continuously enhanced. It is because of this that the application of image recognition technology in digital twinning technology becomes a reality. However, the key technology of digital twin, the virtual-actual mutual control technology, is not mature enough, and the image recognition technology applied in digital twin technology also has the problem of coordinate system transformation, which becomes an important link of image recognition technology in digital twin. Methods: Based on the above two problems, we take the tangram splicing project as an example to realize a virtual-actual mutual control method, so that the digital twinning technology can be well presented. Furthermore, we implement an image recognition applied to the conversion of digital twinning technology, so that digital twinning technology and image recognition have a seamless connection, allowing the application range of digital twinning technology to be further expanded. Results: In this paper, image recognition technology is successfully applied to digital twin technology by adopting the conversion between different coordinate systems and the real and virtual real time control, which makes the applicability of digital twin technology in high-tech fields such as smart factories and smart manufacturing to a higher level. Conclusions: Finally, through the tangram splicing project, the motion trajectories of the robotic arm and the tangram are consistent with the virtual robotic arm and the tangram in the computer. It is proved that our method can well combine digital twinning technology and image recognition technology.

As the core of artificial intelligence, image recognition technology has been widely used in various fields and has brought great convenience to modern life. This paper expounds the image recognition technology and its principle, analyzes the process of image recognition based on artificial intelligence algorithm, and studies the application path of image recognition technology based on artificial intelligence algorithm. Finally, the advantages of artificial intelligence image recognition and its future development are pointed out.

In order to improve the defect recognition efficiency of transmission lines, the industry is currently using aerial images for automatic visual defect detection to ensure the safe operation of transmission lines. This paper proposes a method for defect recognition from coarse to fine, based on convolutional neural networks and connected domain algorithms, to improve recognition accuracy. The recognition speed is improved by using the knowledge distillation method of target detection networks based on decoupled features, adversarial features, and attention features. It has been found that the optimized recognition model improves the precision rate by 7%, the recall rate by 8%, and the average accuracy rate by 10%. The FPS of the model optimized by knowledge distillation is 62.5, and the average value of the FPS of other versions of this model is 47.35. It is believed that the two optimization ideas introduced in this paper can enhance the previous transmission line defect recognition algorithm in terms of accuracy and recognition speed.

In order to be able to make the image recognition and color recognition better fusion and application, this paper uses the YOLO v2 network that achieves excellent results in the target detection field, which can be through the training of traffic light samples to achieve the high speed and accuracy of identifying and positioning the traffic lights, in combination with the HSV color model, design the ratio of the red and green colors in the located area, identify the red and green colors, and then determine the status of the traffic lights. Through this study, we can understand the possibility of using YOLO v2 network in more fields, and at the same time combine color recognition to make YOLO v2 can get more extensive application.

with the continuous progress of the society, in recent years, the image recognition technology has a rapid development, whose new application fields are expanding. The intelligent automotive electronic information system and traffic monitoring research based on image recognition have become a new means to solve some social problems. At present, cars have become a necessity for many families, but at the same time, it has also brought a series of social problems, for example, the elderly and children were locked in the car and suffocated by the tragedy is often reported. In order to prevent these accidents, this paper proposes an intelligent vehicle-mounted/maintenance alarm system based on image recognition. Using deep learning algorithm of system main control board as the carrier, the system connects the temperature, pressure and CO2 concentration sensor and camera equipment, with the aid of technologies such as 4G communication technology, image recognition, automatically monitoring in the car, when there were something abnormal, it would timely alarm to remind the owner, and start the ventilation system of emergency handling, in order to solve corresponding problems.

Since insulators play an essential part in transmission lines, insulator defect detection could be an important assignments for intelligent inspection of high-voltage transmission lines. In this paper, an improved YOLOv5 algorithm for insulator defect detection task for aerial images with various backgrounds is proposed. We use collected aerial images of insulators with one or more defects in different scenarios, perform data augmentation of exposure and noise on the images to expand the sample (expanded to 2125 images), and establish a dataset by combining aerial images of normal working insulators. By adjusting the CSP(Cross-Stage-partial-connections) and CBL(Convalution-BatchNorm-Leaky_relu) modules in the YOLOv5 model to change the depth and width of the model, change model parameters, and build five different scale YOLOv5 models to further meet the real-time task. ResNet and DarkNet are used for the transfer learning of the YOLOv5 model, and various optimization methods are used in the Backbone structure, Neck structure and output of the model, then the established data set is trained and tested on each YOLOv5 model. Among them, the YOLOv5n model has the fastest detection speed, which can reach 10ms, and the precision also reaches 95%. The YOLOv5x model has the highest precision, reaching 97%, and the detection speed is 21ms. These models are all able to satisfy the accuracy and real-time mission in the process of aerial photography and analysis among which YOLOv5n can achieve lightweight tasks while being efficient enough.

Abstract: The paper presents sharing several of experiences and practices on smart robotic application for overhead transmission line maintenance and inspection. The robotic puller passes over barriers below the power line, such as the road with traffic, power distribution lines, river, or vegetation making tasks achieved conveniently, safely, and rapidly without impact on nearby communities.The target of the research is to create a mobile robot prototype for inspection of overhead power lines. The inspection robot shall crawl along the ground wire and transpose autonomously across installed equipment on the ground wire, such as vibration dampers, suspension clamps, compression dead ends, etc. In addition, the inspection robot is able to take photos and videos during a transmission line inspection . Using the robot, transmission line inspection’s labor cost can be reduced, and the new method helps improve patrol and inspection efficiency, comparing to the conventional manpower method. Trough utilization of the new maintenance and inspection robots, utilities can minimiz transmission line operation & maintenance budget.

With the rapid development of information science and technology, image recognition technology plays an increasingly important role in the field of information security. However, in practical application, image recognition is easily affected by factors such as illumination, occlusion, background and other non-ideal conditions, so it is of great practical significance to seek robust image recognition technology. Sparse representation and collaborative representation can capture the essential features of face image, and obtain better recognition effect in image recognition. Therefore, this paper proposes an image recognition algorithm based on information fusion of sparsity and synergy. Experiments are carried out on the problems of collaborative representation classification and single sample image recognition. Experimental results show that, compared with sparse representation classification, collaborative representation classification achieves higher classification accuracy. When part of the pixel value image is occluded by 10%, the recognition rate of sparse representation algorithm is 99.1%, and the recognition rate is very good. Both algorithms have achieved very good recognition results in image recognition. Experiments show that sparse representation algorithm and collaborative representation algorithm improve the recognition rate of images.

Image recognition technology is an important part of machine vision applications. In the process of image recognition, the perception mechanism must exclude the input redundant information and extract important information. Loess is rich in calcium carbonate, and its particles are mainly composed of silt soil with obvious pores between soil particles. Loess in the basin is characterized by high water content, easy collapse, easy collapse, and strong permeability. Loess is very vulnerable to the influence of water, which greatly reduces the support of loess. Most of the quality problems of loess foundation-related projects in the basin are caused by insufficient support. In this paper, the loess in the watershed is selected as the investigation object. Firstly, in order to investigate the change of mechanical properties of loess, freezing thawing cycle, alternation of drying and wetting, and non-direct shear and compaction tests were carried out under these combined effects. Then, combined with the numerical simulation software geo-studio, the effects of temperature field, infiltration field, freezing and melting cycle, and wetting and drying alternation on the slope stability were simulated. Aerobic exercise is helpful to maintain the physiological health of the subjects; the results show that aerobic exercise with small intensity, moderate intensity, and medium intensity has a positive effect on the inhibition and conversion function of athletes' brain; aerobic exercise of three intensities can promote the development of brain refresh function; the results show that aerobic exercise of small plus medium intensity and small plus intensity has a positive effect on the brain activation of switching and inhibition function when athletes complete cognitive tasks. Based on the image recognition technology, this paper discusses the mechanical properties of loess in the basin and studies the monitoring of aerobic exercise data of athletes.

Image recognition technology is an important branch of artificial intelligence research, using the computer to process, analyze and understand the image, in order to identify different patterns of objects. Image recognition technology is currently used in a wide range of applications, such as face recognition, fingerprint recognition, terrain survey, license plate recognition, etc. However, due to the possible existence of multiple categories in images, the blurring of graphic boundaries affects the results of image recognition. In this paper, we propose a multi-category multi-task image recognition method based on deep metric learning (MMDML). Specially, we combine triplet loss function and softmax loss function to construct the loss function, and take ResNet-50 network training to fit the optimal loss function for image recognition. To demonstrate the effectiveness of the proposed method, our method is compared with the other two methods on three common image recognition datasets, namely ImageNet, PASCALVOC, and Caltech. And the experimental results show that our algorithm has the highest Rank-1 and mAP on three datasets.

With the exponential improvement of the integration capabilities of electronic hardware devices, digital images have become an indispensable information carrier, thus promoting the development of image recognition, detection, or tracking technologies. At present, there is a blank in the field of image recognition measured by the action standard of yoga standing three-dimensional movement, which will bring huge room for researchers to play. Based on the traditional image recognition technology, there are cases where missed detection or false detection cannot be correctly identified and made a correct judgment. In this paper, the ability to capture the action is improved by optimizing the network for the positioning of the stereoscopic action target of the yoga station. The linear discriminant analysis (LDA) after adding optimization is used to reduce the dimensionality of the captured image data, which is beneficial to improving image recognition rate and reducing image loss rate. Through the analysis of the three-dimensional movement of the yoga station, the model based on the improved algorithm is compared with the traditional model, and the final image recognition accuracy is improved compared with the network model before the improvement. The image recognition error rate steadily tends to 5%, and the loss rate is also below 2%. Through the optimized convolutional neural network, this paper can accurately capture the image position, and the recognition rate has also been greatly improved, which can provide a reference for future research in other fields.