Pipeline defect intelligent detection method based on image processing and deep learning, and application thereof

Abstract

The invention discloses a pipeline defect intelligent detection method based on image processing and deep learning, and application thereof. The intelligent detection method comprises the following steps: carrying out preprocessing including edge detection on pipeline image data, inputting the preprocessed image into a trained detection model, and carrying out defect detection, wherein the detection model is constructed based on a Darknet53 backbone network and a multi-scale feature extraction network. According to the detection method, different defects of the pipeline can be accurately recognized and detected, the model meeting the industrial application level is obtained, the detection efficiency is high, and the environment universality is high.

Description

technical field [0001] The invention relates to the technical fields of Internet of Things and artificial intelligence. Background technique [0002] Pipelines are commonly used liquid and gas transmission media in the petroleum field. Due to the complexity of the working environment and transmission materials, they are prone to corrosion, blockage and even rupture during use. Therefore, the pipelines are regularly inspected to ensure their service life. And security is a must. The traditional manual detection method for pipeline defects is time-consuming and prone to false detection and missed detection. Therefore, an intelligent defect detection method is a better choice. [0003] The existing intelligent detection methods for oil pipeline defects involve computer vision technology and embedded hardware technology. In terms of hardware, cameras and hardware platforms that can sense the environment, analyze scenes, and make corresponding responses are required. However, d...

AI Technical Summary

Problems solved by technology

This method can quickly and accurately identify crack defects, but it is also difficult to apply to the detection of corrosion defects. At the same time, it requires special application scenarios, and the detection efficiency may be poor in ordinary scenarios.

[0008] It can be seen that the existing steel pipe defect detection systems and methods still have many deficiencies, including the direct deep learning algorithm is difficult to distinguish steel pipe defects, and its target detection method requires a high-performance computing platform and a large operating memory. , the direct defect image has low resolution, poor detection and training effects, simple image geometric modeling cannot realize the detection of corrosive defects, and the existing image acquisition methods do not have environmental universality, etc.

Images