Thermal imaging has become a valuable method for quality control, damage detection and the identification of energy losses. As in every other method, the resolution is crucial for the outcome of analyses and experiments. State of the art thermography cameras have sensors with 1024 \(\times \) 768 pixels, which can record thermal images up to 3.1 megapixels via sensor motion and interpolation. Being mobile devices, these cameras cost several ten thousand euros. In this article, we use a comparably cheap 640 \(\times \) 512 pixel thermography camera for stationary applications, in combination with a panorama head, to create geometrically corrected thermal images of around 50 megapixels with full radiometric information. The panorama head is an optical pivotal point adapter, which can rotate a mounted camera to previously defined angular positions and record images automatically. These single images get geometrically corrected and combined to a whole picture. The technique presented in this article can be applied to almost every thermography camera and is therefore flexible in respect to particular applications or experimental questions. With a novel method of data treatment, the complete radiometric information can be maintained, allowing for a better analysis and processing in thermographic software. With this technique, even full 360\(^\circ \) \(\times \) 180\(^\circ \) thermal image panoramas are possible, which can be used for the inventory of buildings. Due to the geometric calibration, these pictures can be combined with dimensionally stable measurements such as e.g. 3D point clouds from laser scanning.

热成像已成为用于质量控制，损坏检测和能量损失识别的有价值的方法。与其他方法一样，分辨率对于分析和实验的结果至关重要。最先进的热像仪的传感器具有1024  \（\ times \）  768像素，通过传感器的运动和插值可以记录高达3.1百万像素的热图像。作为移动设备，这些相机的价格为几万欧元。在本文中，我们使用相对便宜的640  \（\ times \） 用于固定应用的512像素热像仪相机，与全景云台结合使用，可创建具有完整辐射信息的约50兆像素的几何校正热图像。全景头是一个光学枢轴点适配器，可以将安装的摄像机旋转到先前定义的角度位置并自动记录图像。这些单幅图像经过几何校正，并组合成一整个图片。本文介绍的技术可以应用于几乎所有的热像仪，因此对于特定的应用或实验问题具有灵活性。使用新颖的数据处理方法，可以保留完整的辐射信息，从而可以在热成像软件中进行更好的分析和处理。有了这项技术，即使是全360度\（^ \ circ \） \（\ times \）  180 \（^ \ circ \）热图像全景图是可能的，可用于建筑物清单。由于进行了几何校准，因此可以将这些图片与尺寸稳定的测量值（例如来自激光扫描的3D点云）组合在一起。