可见光与热像融合的三维温度模型重建

热像仪因其能够通过检测物体表面热辐射而产生温度图像,近年来被广泛应用于多种工业场合.然而,由于热像图缺少直观的几何信息,当物体表面温度相近时,难以通过人眼分辨物体特征差异.为了解决这个问题,提出了一种结合可见光几何信息与热像图温度信息的三维模型重建方法.首先使用自制标定板进行可见光相机与热像仪的校正;随后,通过运动恢复...

3D Temperature Model Reconstruction Based on Fusion of Visible and Thermal Images

Thermal cameras, which can detect thermal emissions and produce images of radiation, are widely used in various industrial applications in recent years. However, without intuitive geometric information, it is difficult to identify objects on thermal images when adjacent sampling areas have the same temperature. To solve the above problem, a method to reconstruct a 3D thermal model by fusing visual geometry and temperature information is proposed. Firstly, a calibration board is used to calibrate the visible camera and thermal imager. Then, based on camera poses acquired by structure from motion, semi-global matching is used to obtain a coarse estimate of depth maps which are optimized by a refinement method combining geometry and shading term. Finally, with the help of camera parameters and calibration information, mutual information is used to achieve the registration between the visible and the thermal image, and the 3D temperature model is reconstructed through the depth maps.