研究目的
To design an automatic reading system for analog instruments in power plants that overcomes limitations such as high cost, strict camera angle restrictions, and unsuitability for instruments with dense scales.
研究成果
The developed system effectively reads analog instruments in power plants with high accuracy, overcoming previous limitations related to cost, camera angle, and dense scales. It performs better than human readings and previous systems, offering a practical solution for automatic monitoring.
研究不足
The system's performance is dependent on the quality of the captured images and the accuracy of the robot's positioning. Environmental factors such as lighting and instrument condition may also affect reading accuracy.
1：Experimental Design and Method Selection:
The system combines a mobile inspection robot with an image capture device and image processing methods to reduce monitoring costs. Hough transform and perspective transform are used to correct geometric distortion in images. Polar transform is applied to handle dense scales.
2：Sample Selection and Data Sources:
Images of analog instruments in power plants are captured using an image capture device mounted on an inspection robot.
3：List of Experimental Equipment and Materials:
Inspection robot with an image capture device, NVIDIA Tegra TK1 for processing, STM32 F407 MCU for motor driving, and a camera manufactured by HikVision.
4：Experimental Procedures and Operational Workflow:
The robot moves to a predetermined position, captures images of the instrument, processes the images to correct distortion and extract readings, and uploads results to a monitoring terminal.
5：Data Analysis Methods:
The system uses computer vision techniques to analyze images, including Hough transform for line detection, perspective transform for distortion correction, and polar transform for reading dense scales.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
