Enhanced Detection of Insulator Defects Through Environmental Temperature Curve Analysis

Research has pointed out that traditional detection methods often overlook the impact of environmental factors on infrared images, leading to detection errors and misjudgments [1]. To solve this problem, this study introduces a method for detecting abnormal states in insulators by incorporating the environmental temperature rise curve. Initially, the insulator skeleton is extracted from the infrared image using image processing techniques, simplifying the representation of the insulator. Subsequently, the insulator temperature rise curve is derived based on the skeleton temperature. Through a comprehensive similarity analysis between the insulator and environmental temperature rise curves, intricate characteristics of the insulators such as heat transfer and thermal stability are meticulously examined. Ultimately, the method achieves a recognition accuracy of 99.4% on a self-constructed dataset, demonstrating its effectiveness and meeting the practical application requirements.