Research On Infrared Image Enhancement Algorithm For Sea Surface Target Scene

With the development of maritime transportation and tourism in China in recent years,the maritime distress accident occurs frequently.As we all know,timely and effectively maritime search and rescue work can minimize accident loss and ensure the safety of people's life and property.Maritime distress accident occurs mostly at night or under bad weather conditions,such as heavy fog,and SAR helicopters mostly depend on the human eye or telescope to search distress target which is time consuming,laborious and inefficient.Compared with other methods,infrared imaging technology has several advantages,such as ability to penetrate through the smoke,strong ability of anti-interference and the ability to work all day.In infrared imaging system,data acquisition can be achieved through detecting the thermal infrared emitting by object to meet the all weather work requirements of the detection system.However,due to the detector materials,imaging mechanism and the influence of environmental factors,infrared images often have low contrast,fuzzy details and other shortcomings,those all will have serious impact on the follow-up work.This topic is supported by the National Science and Technology Support Project"deep sea surface distress target airborne infrared search and location technology".Because of the low contrast of sea surface infrared image,it is impossible to get any obvious target contour from the original image.Based on the study of the commonly used algorithm of image contrast enhancement,we use high dynamic range compression framework based on bilateral filtering to enhance image contrast.Firstly,the image is divided into base sub-image and detail sub-image by guided filter;secondly,we use bi-histogram equalization based on maximum entropy method to process base sub-image,and use Gamma transform to detail sub-image to enhance detail.Due to bad weather,the image still remains fuzzy after contrast enhancement.On the basis of deep understanding of atmospheric scattering model,we realize that the fuzzy can be removed by using the method based on this model.In this paper,the method estimates transmission by defining the cost function.We use guided filter to optimize transmission to eliminate artificial traces;then,through the atmospheric scattering model we get enhanced infrared image.Finally,in order to verify the effectiveness of the algorithm,we use several image quality evaluation indexes to evaluate the quality of image after enhancement.From the results of visual effect of image and image quality evaluation index,we can conclude that the algorithm adopted by this article can effectively improve the contrast of infrared image and effectively reduce fuzzy caused by the path and bad weather conditions on the target infrared image,and obtain the suitable infrared image for subsequent small target detection.