CAJ | 학술논문

红外热图像测温过程中存在的大量噪声严重制约了红外热图像的精度和准确性，为了准确提取红外热图像中被测物体的相关参数，提出了一种红外热图像二维无网格消噪方法。将双边滤波、异性消噪、无网格消噪进行结合，建立红外热图像二维无网格消噪模型，利用鲁棒回归算法求解模型中相关参数，实现了红外热图像的无网格消噪，提高了红外热图像测温的精度。最后，通过仿真实验验证了提出方法的可行性，为军事、科研等领域的红外热图像的测温提供了理论基础。
홍외열도상측온과정중존재적대량조성엄중제약료홍외열도상적정도화준학성，위료준학제취홍외열도상중피측물체적상관삼수，제출료일충홍외열도상이유무망격소조방법。장쌍변려파、이성소조、무망격소조진행결합，건립홍외열도상이유무망격소조모형，이용로봉회귀산법구해모형중상관삼수，실현료홍외열도상적무망격소조，제고료홍외열도상측온적정도。최후，통과방진실험험증료제출방법적가행성，위군사、과연등영역적홍외열도상적측온제공료이론기출。
The temperature measurement process of infrared thermal image has a large amount of noise,which has seriously hampered the precision and accuracy of infrared thermal image. In order to accurately detect the correlative parameters of the measured object in the infrared thermal image,a 2D meshless denosing method based on infrared thermal images is proposed. Bilateral filtering,anisotropic denoising,meshless denoising are combined to create a 2D meshless denoising model. The model parameters are estimated by using robust regression algorithm,and realize infrared thermal images with meshless denoising. And it enhances the temperature measurement precision of infrared thermal image. Finally,the feasibility of the proposed method is verified by simulation ezperiments and the method lays the theoretical foundation for military,scientific research and other areas of temperature measurement of infrared thermal image.