| Infrared technology is not only widely used in military,national defense,fire safety and other fields,but also widely used in many civil fields.Infrared digital image has some problems,such as blurred details,poor contrast,high noise and dark image.In order to improve the quality of infrared image,we can start with software and hardware.Firstly,in software,we use computer tools to simulate the infrared image algorithms that need to be processed,and compare the advantages and disadvantages of various algorithms.After repeated experiments,we find the algorithm with good results,and finally transplant the infrared image processing algorithm to hardware implementation.On the other hand,by constantly improving the development and production technology level of infrared sensor components,improving the accuracy of hardware products,and then improving the overall quality of products.However,because the development of semiconductor materials technology in China is not mature abroad,there is a certain gap between the research and production technology of infrared detector focal plane technology and foreign countries.Moreover,the research results of semiconductor materials and devices are not obvious,so the research of algorithm is particularly important.Infrared focal plane plays a key role in the whole infrared industry.In this paper,the algorithm is used to enhance the details of infrared image.Through extensive reading of the infrared image literature,it is found that the infrared image detail enhancement processing method has always been the pain point of the infrared technology.This paper combines the spatial stratification technology and histogram equalization method in the infrared digital image detail enhancement algorithm,uses the BEEPS algorithm filter to extract the infrared digital image from the fundamental frequency layer,which is the energy layer,and the fundamental frequency layer equalizes with the histogram.Technical processing,so that the pixels of the whole basic frequency image can occupy the whole gray level as far as possible and evenly distribute,so that the infrared digital image will have a high contrast appearance and show a greater change in gray tone.After the difference between the original infrared image and the base-frequency image,the detail layer,i.e.the high-frequency layer,can be obtained,and then the detail layer is enhanced.The enhanced detail layer and the base-frequency layer after histogram equalization are fused.When processing the infrared image,the effect of the final infrared image can be adjusted by the gain of the detail layer and the parameters of the BEEPS filter to make the infrared image more effective.The effect of the image is the best,and the infrared image after the final enhancement details is obtained.Considering from subjective vision,the details of infrared image are more obvious,the whole infrared image is brighter,and the contrast of infrared image is improved.Infrared image detail enhancement algorithms generally include the following types: 1)histogram type 2)frequency domain transformation type 3)spatial hierarchical processing type;these methods are common,but the processing effect is relatively simple,often can be used as a preprocessing tool,in real application,these methods can be combined with the details of infrared digital images.This paper combines the basic methods of BEEPS algorithm filter and infrared image detail enhancement algorithm: spatial hierarchical processing and histogram equalization.The BEEPS algorithm is proposed by Philippe Th venaz,Daniel Sage and Michael Unser,members of the IEEE Association,in 2012.This paper first introduces the BEEPS algorithm to extract the details of infrared digital images.It is found that the calculation of BEEPS algorithm is only related to the size of the image itself,but not to the data,filtering parameters and final smoothness of the image.Compared with the current main infrared digital image detail processing methods,bilateral filters,joint bilateral filters and guide filters,bilateral filters also adjust the data locally and fix someof them after derivation.When parameters are used,the formula of bilateral filter and bilateral filter can be converted to each other,and when using MATLAB tool,the processing time of infrared image can be saved by one order of magnitude compared with bilateral filter.The research status and development trend of infrared digital image details are analyzed.According to infrared image characteristics and existing infrared digital image details algorithm,BEEPS algorithm is proposed to extract infrared image details.The realization of this paper saves a lot of memory and time,and will produce enormous economic benefits.When using background variance-detail variance(BV-DV)index and root mean square contrast(RMSC)index to evaluate the performance of this method,it is found that BEEPS algorithm is better than current bilateral filters,pilot filters,joint bilateral filters and other algorithms,and the filter details of BEEPS algorithm are better. |
PDF Full Text Request