SAR Image Processing for Vessel Detection Near Land
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing marine surveillance methods using synthetic aperture radar (SAR) images struggle to accurately detect vessels near land due to interference from land objects, such as bridges or cranes, which complicates setting luminance thresholds for binarization, leading to inaccurate detection of vessels and water areas.
Innovation Solution
An image processing method that generates a difference image between a focused SAR image and a background image, using geographic information to set a water area and determine luminance thresholds for binarization, allowing for the suppression of land object interference and accurate detection of vessels.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If luminance threshold binarization is used to detect vessels in SAR images, then detection process is simple, but detection accuracy deteriorates when land objects with high luminance interfere with vessel detection
Solution Approach 1:
The patent divides the SAR image into multiple regions based on geographic information: water areas, land areas, and extended land areas. By segmenting the image and applying different processing rules to different regions, the method maintains simple binarization processing while improving vessel detection accuracy near land by excluding land objects from vessel candidate detection.
Solution Approach 2:
The patent applies different luminance threshold values for binarization based on the local region characteristics. Water areas use one threshold while land areas and extended land areas use different thresholds or are excluded from vessel detection. This local differentiation allows simple processing to maintain high accuracy in different regions.
2Measurement precision
If luminance threshold is lowered to detect low-luminance vessels, then more vessels are detected, but false detection of land objects increases
Solution Approach 1:
The patent extracts and excludes land objects from the detection process by defining extended land areas that extend from actual land boundaries into water areas. By removing these regions from vessel candidate detection, the method can use lower luminance thresholds to detect faint vessels without falsely detecting land objects as vessels.
Solution Approach 2:
The patent performs preliminary region classification using geographic information before the actual vessel detection process. By pre-defining water areas, land areas, and extended land areas, the method prepares the detection framework in advance, allowing sensitive vessel detection in water areas while preventing false detections in land-adjacent regions.
3Object-generated harmful factors
If luminance threshold is raised to avoid detecting land objects, then false detection decreases, but vessel detection accuracy deteriorates
Solution Approach 1:
The patent applies different luminance threshold values for binarization based on the local region characteristics. Water areas use one threshold while land areas and extended land areas use different thresholds or are excluded from vessel detection. This local differentiation allows simple processing to maintain high accuracy in different regions.
4Measurement precision
If geographic information is integrated to define water areas and set thresholds, then detection accuracy improves, but processing complexity increases
Solution Approach 1:
The patent uses geographic information (land boundary data) as an intermediary to guide the image processing. By overlaying geographic information on the SAR image to define water areas, land areas, and extended land areas, the method achieves high detection accuracy without requiring complex image analysis algorithms. The geographic information acts as a mediator that simplifies the overall processing complexity.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image processing device 100 of the present invention includes an image generation means 121 for generating a difference image representing a difference between an object image that is an image including an area from which a mobile body is to be detected and a corresponding image that is another image including an area corresponding to the area of the object image, and a detection means 122 for detecting the mobile body from the object image on the basis of the object image, the corresponding image, and the difference image.