Research On Preprocessing Technology And Target Detection Method Of Marine Infrared Images

Sea target detection has important application value in marine safety construction and maritime emergency search and rescue.Because of the advantages of day and night monitoring and strong fog permeability,infrared imaging technology plays an irreplaceable role in the detection of sea targets.However,due to the influence of natural conditions and detectors,infrared images are prone to blur and non-uniformity,which makes the detection of infrared targets in complex environments facing great difficulties and challenges.In order to better to extract target from infrared image in sea-sky background,the following research work has been carried out for three problems in the paper: blocking artifact removal,sea-sky line extraction and target detection in sea surface infrared images.1)A blocking artifact removal method based on iterative constrained optimization is proposed to solve the problem of blocking artifact in low bit rate marine infrared images.Firstly,according to the cause of blocking artifact,a measurement method is proposed to measure the extent of blocking artifact.Then,a constrained optimization objective function is constructed based on DC component compensation model.Then constrained optimization iteration algorithm is used to obtain the optimal compensation value of image blocks.Finally,combined with post-processing method the blocking artifact in infrared image is removed.The experimental results show that the proposed method can effectively reduce the gray difference between image blocks,better remove the blocking artifact in infrared images,improve the image quality,and be conducive to the subsequent sea-sky line detection and target detection.2)Aiming at the problem of sea-sky line detection in sea-sky background,a sea-sky line detection method based on boundary priori double model Bayesian decision is proposed.Firstly,two Bayesian decision models are constructed based on the boundary prior information.Then,the sea and sky regions are segmented by rough and fine partition of sea and sky blocks.Finally,the sea-sky line candidates are fitted by the RANSAC technology to realize the detection.The experimental results show that the proposed method can detect the sea-sky line under the complex sea clutter interference quickly and effectively,and improve the accuracy and speed of the follow-up target detection.3)Aiming at the problem of target detection in sea surface infrared image,a method of target detection based on sea water background in discrete cosine domain is proposed.Firstly,the background area of pure sea water is extracted from a single frame image by using fuzzy C-means method.Then,a Gauss background suppression filter is designed using pure sea water background.Finally,Target extraction is realized by filtering the background area of non-pure sea water in frequency domain.The experimental results show that the proposed method can accurately and effectively detect the infrared target in the sea background,and has a good background suppression effect.