Research On The Registration And Fusionalgorithm Of Visible And Infrared Image

At present,infrared thermal imager are widely used in many fields,such as the detection of targets in the military field,the detection of electrical equipment failures in the industrial environment,the inspection and repair of internal circuits of electronic products,the inspection and maintenance of building pipeline heating equipment,and fire alarms,etc.,and The drone also involves image fusion technology.Especially since the emergence of the Covid-19 last year,infrared thermal imager cameras have been used to judge and detect fever groups at the entrances of population flow,such as shopping malls,train stations,and office buildings.Greatly improve the efficiency of epidemic prevention and reduce manual workload.The fusion technology of visible image and infrared image,as the core technology of image processing in infrared thermal imager,is directly related to the performance of infrared thermal imager.Research on the fusion technology of visible image and infrared image can further improve the performance of thermal imager.The performance makes the infrared thermal imager more comprehensive and effective for the acquisition of scene information,and better combines the useful information complementary to the infrared and visible images,which is convenient for the observer to judge the scene,and reduces the requirement of infrared thermal imager to single camera performance requirements.In order to further improve the image fusion performance,this paper conducts algorithm research on the registration and fusion of visible image and infrared image for infrared thermal imager,and designs an image algorithm suitable for infrared thermal imager.This paper carries out the following image algorithm research under this technical requirement and background conditions.First of all,due to the influence of the working environment of the infrared thermal imager and internal hardware reasons,it is necessary to pre-process the collected original images,and the appropriate pre-processing method should be studied around the infrared thermal imager to perform the visible image and infrared image.Preprocessing operations include conversion of infrared grayscale data,image de-manipulation,image enhancement,image interpolation and resampling,and pseudo-color processing of infrared images.Secondly,there is a certain distance between the infrared camera and the visible camera in the infrared thermal imager,and the imaging resolution of the two cameras is different,resulting in a certain resolution of the source image collected by the two cameras and the shooting angle of the image.It is difficult to guarantee the effect of the fused image if it is directly used for fusion.In this paper,the dual camera calibration is used to calibrate the infrared camera and the visible camera.The dual camera calibration registers the visible image with the infrared image by obtaining the internal and external parameters of the infrared camera and the visible camera and the relative relationship between the two cameras.Then researched and analyzed the commonly used algorithms for image fusion,and selected the weighted coefficient fusion algorithm for infrared thermal imagers to fuse infrared images and visible images.Experiments have also proved that it is suitable for infrared thermal imagers,the fusion effect also meets the needs of infrared thermal imager.In order to evaluate the fusion image,the subjective evaluation method and the objective evaluation method are studied.Because the subjective evaluation method is relatively affected by external factors,the objective evaluation method evaluates the image by quantifying the mathematical model in the image,which mainly depends on the image itself,it has been verified through experiments that the image fusion algorithm with weighted coefficients used meets the product requirements of infrared thermal imagers.Finally,through experiments in the OpenCV-Python experimental environment and the development environment of the lower computer,the original image collected by the infrared thermal imager can effectively deal with the problems in the original image through the image preprocessing method designed in this paper,so as to ensure the image after preprocessing meets the requirements of image registration.The dual camera calibration method designed in this paper can register visible images and infrared images,and meets the requirements of image fusion.Finally,the weighted fusion algorithm is used to fuse the visible image and the infrared image.The fusion image evaluation proves that the fusion algorithm in this paper meets the requirements.Infrared thermal imager needs,the details and information contained in the fusion image are richer,the fusion image is more understandable,the information obtained by the observer is more comprehensive and accurate,the occurrence of false judgments is reduced,and the robustness and fault tolerance of the system is improved.