Telegraph pole sign target detection and identification method based on end-to-end regression model

Abstract

The invention discloses a telegraph pole label target detection and identification method based on an end-to-end regression model, and the method comprises the following steps: making a diversified data set of telegraph pole label images, and carrying out the training of a telegraph pole label for an end-to-end one-time regression target detection model under a classification network framework; extracting position coordinates of the label in the image, and performing character segmentation by combining prior information and a projection analysis method; simplifying the training of a characteridentification network by adopting transfer learning so as to improve the overall efficiency; and finally, selecting an HSI color space to identify the background color of the sign so as to reduce theworkload of image analysis. According to the invention, an efficient end-to-end target detection method is adopted to realize accurate identification of effective information of the telegraph pole sign, and the efficiency of routing inspection path planning and rural power grid reconstruction investigation can be improved to a certain extent.

Description

technical field [0001] The invention belongs to the technical field of electric pole signage image recognition in power transmission line inspection work, and specifically relates to the end-to-end target detection of electric pole signage based on a one-time regression model, the use of transfer learning to simplify the training process of character recognition networks and the ability to simplify A method for background color recognition in HSI color space for image analysis work. Background technique [0002] my country's power grid construction is developing rapidly, and social production and life are increasingly dependent on the power system. With the improvement of government and national awareness of environmental protection, the proportion of power consumption in many energy-consuming fields is also increasing. Reliable operation places ever higher demands. Moreover, with the construction of the ubiquitous power Internet of Things, all aspects of power system operat...

AI Technical Summary

Problems solved by technology

[0004] In view of the low recognition rate of the traditional image recognition method for utility pole signs and the time-consuming and labor-intensive manual recognition method, the present invention provides a utility pole based on an end-to-end regression model Sign target detection and recognition method to improve the overall recognition efficiency

Images