| Temperature is one of the most basic physical quantities in nature,and plays a very important role in production and life.This article introduces the research background of this study,the development history and research status of high temperature measurement,points out the limitations of the current various high temperature measurement methods,and at the same time,discusses the types and status of high temperature measurement methods based on digital image processing.Due to the problems of insufficient accuracy,narrow temperature range and complicated structure of the equipment,the research goal of visible and near-infrared imaging temperature measurement technology is proposed.The target temperature measurement accuracy is set to a relative error of not more than 5%,and the dynamic range is 1000-3000 K.This article introduces the basic principles of radiation temperature measurement in detail,analyzes the factors that affect the accuracy of radiation temperature measurement,simulates and optimizes the temperature measurement system,designs experiments based on the optimization results,builds a platform,and verifies the feasibility of the program.The specific work is as follows:(1)Establish a visible and near-infrared imaging temperature measurement system structure model,specify the factors that may cause temperature measurement errors in the model,and for these factors,simulate and optimize the temperature measurement operating wavelength and working narrowband bandwidth.First,the RBF neural network is used to fit the response curve of the selected visible and near-infrared detector,the fitting result is combined with the principle of colorimetric temperature measurement,and the particle swarm algorithm is used to synthesize the calculation results and analyze the influence of emissivity on the working narrowband bandwidth,And comprehensively consider the actual production process and shooting distance and other factors,and finally determined the working wavelength is 650 nm and 800 nm,and two working narrowband bandwidth 10 nm and 20 nm.(2)According to the simulation modeling optimization results,design the experimental process,build the experimental platform under the existing experimental conditions,select the wide-spectrum low-illumination CMOS camera RS-A1304,and use standard high-temperature blackbody radiation sources to capture the collected images Carry out system calibration,and use least squares method to fit the experimental data to correct the system error,and then carry out temperature measurement experiment on the halogen lamp filament to verify the accuracy of the system.(3)From the experimental results,the system has an average relative error of 1.9% when operating at a narrow bandwidth of 10 nm,and an average relative error of 3.0% when operating at a narrow bandwidth of 20 nm,which meets accuracy requirements;at the same time,according to the temperature measurement range of the system,the multi-level adaptive pseudo color coding method suitable for the system is improved and compared with the classic pseudo-color coding method,proved the superiority of the algorithm,enhance the perception of temperature on the vision;Finally,the visible and near-infrared imaging temperature measurement software is designed to realize the functions of image import,pseudo color coding and temperature calculation.And summarize the full text and put forward the future development direction of this study.From the experimental results,the visible and near-infrared two-color imaging temperature measurement system designed in this paper can meet the design requirements of high precision and large dynamic range,and achieve the expected results. |
