激光激发下拉曼光谱相对辐射定标模型研究

为了提高变压器油中气体检测性能,提出基于激光激发下拉曼光谱的相对辐射定标模型,采用激光成像技术进行拉曼光谱图像成像处理,提取拉曼光谱图像的边缘轮廓特征量,采用多尺度小波降噪方法进行激光激发下拉曼光谱图像降噪滤波,在邻域内采用颜色梯度分解方法进行拉曼光谱图像区域融合性增强处理,根据激光激发下拉曼光谱图像RGB值与邻域均值实现拉曼光谱相对辐射标定和检测,提高拉曼光谱图像的相对辐射定标能力。仿真结果表明,采用该方法进行激光激发下拉曼光谱相对辐射定标的准确性较高,识别精度较高,图像信息增强性能较好。

Study on the relative radiometric calibration model of raman spectra under laser excitation

In order to improve the performance of gas detection in transformer oil,a relative radiometric calibration model based on Raman spectrum under laser excitation is proposed.The laser imaging technology is used to process the Raman spectrum image,and the edge contour feature of Raman spectrum image is extracted.The multi-scale wavelet de-noising method is used to de-noising and filtering Raman spectral images under laser excitation,and the color gradient decomposition method is used to enhance the fusion of Raman spectral images in the neighborhood.The relative radiation calibration and detection of Raman spectrum image are realized according to the RGB value of Raman spectrum image and the neighborhood mean value under laser excitation,and the ability of relative radiometric calibration of Raman spectrum image is improved.The simulation results show that the accuracy and recognition accuracy of Raman spectrum relative radiometric calibration under laser excitation are high,and the image information enhancement performance is better.