Research And Application Of Infrared Thermal Image Stitching Technology

The infrared image only reflects the temperature difference between objects,which makes the infrared monitoring technology has the advantages of long-distance?non-contact and free from harsh environment,so it can be widely used in military target detection and early warning,power equipment online monitoring,industrial testing and other fields like this.However,most of the Infrared Thermal Imager's range is narrow and the resolution is not high,it is often difficult to show each parts of the large target object's appearance,which brings a lot of inconvenience to the image analysis and the judgment of temperature trend,while image stitching technology provides a more appropriate solution.We first study the stitching technology of infrared image in this paper,and focus on the feature point extraction and image registration of infrared image,then put forward the corresponding improved algorithm.Then the infrared thermal image acquisition and splicing system is designed,and verified by experiments for its practicability and reliability.In the image preprocessing,we analyze the infrared imaging theory first,and study the denoising and enhancement technique of infrared image.Finally,we analyze the experiment results of simulation experiments of the infrared image.In the study of image registration,three registration methods are introduced and simulated,and feature-based image registration scheme is determined.In feature points extraction,the classical corner extraction algorithm and local feature invariant feature extraction algorithm is studied further,and the simulation and result analysis of programming experiment is carried out.An improved feature point extraction method in multi-scale space is put forward,and validated for its feasibility by programming experiment.In the matching process,the rough matching between the matching images and the improved RANSAC algorithm are studied,and validated for the feasibility of the algorithm by programmingexperiment.In the research of image fusion,this paper introduces several basic fusion methods,which completes the last step of image mosaic,and then do the programming experiment to generate the mosaic diagram of the two graphs.In the design of software system,this paper designs a set of image acquisition and splicing system based on infrared thermal imager according to the previous research results.This system not only can complete the control of the imager,the detection of real-time video's point temperature and the acquisition of the thermal image,but also can select the corresponding algorithm to process and mosaic the image collected by the thermal imager,which brings the following image analysis work more convenience and makes it a useful and meaningful tool for users.