Research On Small Target Detection In Infrared Sea Surface Based On Space-time Multi-features

As infrared technology is widely used in civil and military fields,the detection of infrared small target is becoming more and more important,but the detection of small target on the sea is still one of the difficult problems in the field of signal processing.Because of its own characteristics and background characteristics,small targets on the sea are often not easy to be detected.The main difficulties are: 1)Small targets are often difficult to detect in the background because of their weak characteristics,no texture and shape characteristics;2)Due to weather,sea conditions and other factors,the sea scene often produces complex interference,which makes small targets easily submerged in the complex background,making small target detection extremely difficult.Therefore,the infrared small target detection technology in complex background is an important content in the research of complex long-range targets at sea.In this thesis,the sea-sky-line is used to divide the sea and sky areas to reduce the impact of the sky area on the detection of small targets on the sea surface.Then,the detection of small targets on the sea surface is carried out.The specific research is as follows:Firstly,according to the difference between sea and sky area,the characteristics of sea-sky-line is used to divide the line of two regions.This thesis proposes a sea-sky-line detection algorithm based on instantaneous coefficient of variation(ICOV)based gradient and Hough transform on sea-sky-line detection in view of the interference of weather and sea conditions.The gradient of the original image is calculated by using the instantaneous coefficient of variation based gradient,and then the binary gradient image is obtained by using the adaptive threshold.Then,the sea wave pixels are merged by morphological operation,and then the edge is extracted.Finally,the sea-sky-line is extracted by Hough transform.Experiments show that the algorithm can suppress the influence of sea clutter and fish scales well,and detect the sea-sky-line correctly.Secondly,according to the difference between the sea surface small target and the sea surface background in space domain and time domain,infrared sea surface small target detection algorithm based on space-time multi-characteristics is proposed.The infrared small target is the extreme point in the local area in the airspace,and its singularity makes it different from the background to some extent.The candidate small target can be detected by detecting the background difference of the small target.At the same time,a large number of candidate false small targets can be filtered by multi-feature classification in the spatial domain.The real small target has continuity and consistency in time domain,while the false target has no continuity or consistency in time domain,which can realize the detection of the real small target.Experimental results show that compared with some other methods,the proposed method has good accuracy and robustness in complex infrared sea surface.