Two-dimensional code based substation equipment object recognition method

Abstract

The invention discloses a two-dimensional code based substation equipment object recognition method, which comprises the steps of representing equipment objects of each substation by adopting a mode of two-dimensional code labels and adding equipment information into the two-dimensional code labels, automatic inspection control of video, two-dimensional code label object recognition, two-dimensional code label information recognition, equipment object recognition and video location storage. The method disclosed by the invention is based on two-dimensional code technologies and image recognition technologies of the internet of things, does not need to construct an equipment model or a sample library in advance, does not need manual configuration, avoids manual participation required in other methods, greatly improves the work efficiency, reduces the workload, and can search the equipment objects and acquire equipment names accurately and automatically. In addition, the two-dimensional code labels are high in recognition rate and barely influenced by the outside world, so that the accuracy and the reliability of recognition can be well ensured. Furthermore, types and names of the monitored equipment objects can be judged conveniently according to equipment information contained in the two-dimensional code labels.

Description

technical field [0001] The invention relates to the technical field of electric product accessories, in particular to a two-dimensional code-based substation equipment object recognition method. Background technique [0002] With the gradual promotion and construction of intelligent substations and unattended substations in the power grid, based on the video intelligent analysis technology of the substation video monitoring system, the demand for technical support for the production and operation of the power grid is constantly increasing. System, due to the large number of video surveillance equipment and the large number of production equipment, in the video recognition process, it is necessary to pre-configure the camera preset position manually, and adjust the camera monitoring angle to the specified device object before the surveillance video can be analyzed using video intelligence. Technology is used in the production and operation of the power grid. This method leads...

AI Technical Summary

Problems solved by technology

[0004] However, the method based on template matching technology needs to build the feature model of the equipment object first. Due to the large number of equipment objects and the diversity of equipment types in the substation, a large number of equipment models need to be constructed. Different factors, many occlusions between devices, greater weather influence, different camera resolution, different lighting conditions and other factors make it difficult for the constructed feature model to match in practical applications, and can only be used in the laboratory. In testing and verification, practical application is of little significance

The use of video intelligent recognition technology based on machine learning algorithms depends on the collection of a large number of device model samples. The comprehensiveness and accuracy of the collected model samples will directly affect the accuracy of the device model constructed by machine learning, which will further affect the video quality. The accuracy of recognition requires a large workload, and it is necessary to continuously optimize and adjust the sample library and machine learning algorithms, and since the scene of substation video surveillance usually contains many devices, it will greatly affect the accuracy and reliability of video recognition , at the same time, using the above two methods can only identify the type of equipment, but cannot specifically identify the name of the equipment, it is difficult to meet the needs of equipment status identification during production and operation

Images