Instrument identification and reading method based on lightweight network

Abstract

The invention discloses an instrument identification and reading method based on a lightweight network, and relates to the technical field of instrument measurement, and the method comprises the following steps: obtaining an original image with an instrument; performing instrument target identification and classification on the original image to obtain a classified instrument image; identifying the classified instrument image by adopting a fine-grained classification network combined with mixed attention clipping to obtain range information of the instrument image; based on the measuring range information of the instrument image, obtaining an instrument reading; according to the method, the learning ability of the model for image detail targets is improved through a multi-scale feature fusion means, the operation efficiency of the algorithm is improved through single-scale YOLO calculation, and meanwhile, the detection precision of instrument targets is not reduced; and automatic identification, reading and data storage of the instrument can be realized, tedious manual identification and recording work is omitted for workers, the efficiency of the workers is improved, and meanwhile, the utilization rate of data is improved.

Description

technical field [0001] The invention relates to the technical field of meter measurement, and more specifically relates to a light-weight network-based meter identification and reading method. Background technique [0002] An instrument is a general term for instruments used to measure certain physical quantities. Through some physical or chemical principles, it can intuitively display some physical quantities that are difficult to feel intuitively by means of scales or indications, such as temperature, pressure, humidity, flow and Voltage, etc., are widely used in various aspects such as industry, national defense and people's life. Especially in some key areas of national defense and military industry, the reading of instrument data often requires a lot of labor, and most of these instrument data are currently stored in paper media, which is difficult to further utilize. [0003] At present, most of the industry adopts two methods for instrument identification. The more a...

AI Technical Summary

Problems solved by technology

However, the manual reading method has the following disadvantages: in some extreme environments, such as high temperature, high pressure, and high radiation conditions, the manual reading method is more dangerous; if a large amount of real-time instrument data needs to be obtained, then It requires a lot of manpower, which is inefficient and cost-effective; it is difficult to carry out 24-hour long-term uninterrupted monitoring manually; the readings obtained manually are often stored on paper media, and need to be further transcribed into the computer before they can be used in the next step. Further increase manpower consumption

Therefore, the above two methods for instrument identification cannot meet the current production methods, so it is an urgent problem for those skilled in the art to study an intelligently identifiable instrument identification and reading method

Images