A pumping unit recognition method and system based on target feature training

Abstract

The invention relates to a method and system for identifying a pumping unit based on target feature training, including an image acquisition module for acquiring infrared peripheral vision images and converting them into image input sources for a monitoring identification module; and a monitoring identification module for processing the acquired Infrared image input, detect the pumping unit in the infrared image, and identify the position and working form of the pumping unit based on deep learning; the state analysis module is used to receive the processing results of the monitoring identification module, and continuously analyze the work of the pumping unit The working state of the pumping unit can be obtained. The invention dynamically acquires the position and working form of the pumping unit in the well network by intelligently marking the pumping unit, so that the identification of the working state of the pumping unit is more accurate and better detection effect is obtained.

Description

technical field [0001] The invention relates to the technical field of target detection, in particular to a method and system for identifying an oil pumping unit based on target feature training. Background technique [0002] The detection and identification of the working state of the pumping unit plays an important role in oil field detection. By detecting whether the pumping unit is in working condition, it can effectively detect the failure of the pumping unit itself and the safety status of the surrounding environment when the pumping unit starts and stops. testing to ensure the safety of oilfield production. [0003] The traditional detection mode of the working state of the pumping unit is divided into two types. One is to detect the working state of the pumping unit through the system integrated sensor. The photoelectric observation turntable is used to perform staring detection for each pumping unit. The disadvantage of this mode is that the speed is slow and the p...

AI Technical Summary

Problems solved by technology

[0003] The traditional detection mode of the working state of the pumping unit is divided into two types. One is to detect the working state of the pumping unit through the system integrated sensor. The photoelectric observation turntable is used to perform staring detection on each pumping unit. The disadvantage of this mode is that the speed is slow and the position of the pumping unit needs to be calibrated in advance.

Images