Target recognition method based on fusion of visible light and infrared two-band image target features
A target recognition and target feature technology, applied in image analysis, image data processing, character and pattern recognition, etc., can solve problems such as unfavorable target recognition, poor feature point matching effect, pixel-level fusion image ghosting, etc. Accurate rate and the effect of reducing errors
Active Publication Date: 2020-08-14
江苏鲸天科技有限公司
View PDF6 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] At present, most of the research on multi-source image fusion is based on pixel-level fusion of feature point matching. When the number of different source images increases and the image background is complicated, the feature point matching effect is poor, and there are a large number of mismatched feature point pairs, which eventually leads to pixel There is ghosting phenomenon in the level fusion image, which is not conducive to the target recognition in the later stage
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0027] The method is divided into two parts: visible light and infrared two-band image target registration and visible light and infrared two-band image target recognition. The specific steps are as follows:
[0028] (1) Visible light and infrared two-band image target registration
[0029] When the number of targets is large, the time complexity of the target registration method for visible light and infrared two-band images based on the blind search mechanism is also large, so the present invention proposes a visible light and infrared two-band image target based on first rough classification and then fine registration Registration method:
[0030] ① Image preprocessing, median filtering for visible light images, median filtering and histogram equalization for mid-infrared and far-infrared images;
[0031] ②Study the imaging characteristics of visible light, mid-infrared and far-infrared images, clarify the feature difference between the image target area and the background...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a target identification method fusing target features of visible light and infrared two-waveband images. The method comprises a stage of performing target registration on visible light and intermediate and far-infrared two-waveband images, and a stage of performing target identification on the visible light and intermediate and far-infrared two-waveband images. The method is the target identification method which realizes fusion of the target features of the visible light and infrared two-waveband images on the basis of registering the same target regions of the visiblelight and infrared two-waveband images by taking a target whole region as a starting point.
Description
technical field [0001] The invention relates to a target recognition method for integrating visible light and infrared two-band image target features, and belongs to the technical field of target recognition. Background technique [0002] Modern warfare urgently needs to be able to identify battlefield targets from complex backgrounds around the clock. The thermal radiation wavelengths of important military targets such as aircraft, rockets, and cruise missiles are mostly concentrated in the mid-infrared band of 3-5 microns and the far-infrared band of 8-10 microns; in different atmospheric environments, the mid-infrared band and far-infrared band The transmittance of the band radiation is different. Visible light images have rich color information and texture information, and can capture the details of battlefield targets. Infrared images can be imaged around the clock, and the edge information of the target is well preserved, and the structural information is complete. ...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06T7/33G06T7/62G06K9/46G06K9/62
Inventor 闫钧华谢天夏张寅黄伟蔡旭阳杜君白雪含
Owner 江苏鲸天科技有限公司