Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Small target detection method based on Markov random field and visual contrast mechanism

A small target detection and visual comparison technology, applied in the field of computer vision, can solve the problem of poor detection performance of weak and small targets, and achieve the effect of reducing running time

Active Publication Date: 2019-06-25
绵阳慧视光电技术有限责任公司
View PDF5 Cites 10 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

The above-mentioned single-frame detection methods mostly use a single filter calculation and binarization processing, and the detection performance for weak and small targets is poor.

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 more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Small target detection method based on Markov random field and visual contrast mechanism
  • Small target detection method based on Markov random field and visual contrast mechanism
  • Small target detection method based on Markov random field and visual contrast mechanism

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0084] Such as figure 1 As shown, a small target detection method based on Markov random field and visual contrast mechanism includes the following steps:

[0085] a. Preprocessing the original infrared image of the small target to be identified;

[0086] b. Calculate the saliency map of the preprocessed infrared image based on the Jiugongge algorithm;

[0087] c. Model the inter-frame difference image based on the Markov random field, and segment the motion area and background area;

[0088] d. Fusion the saliency map in the space domain and the segmented area in the time domain, and remove the isolated noise points in the image that are mistaken for the target.

[0089] In this embodiment, the original image is modeled and a saliency map is generated respectively, that is, on the one hand, the original image is firstly subjected to single-frame image preprocessing, and then a saliency map is generated, and finally a region of interest (ROI) is extracted; On the one hand, ...

Embodiment 2

[0091] In this embodiment, on the basis of Embodiment 1, said step a includes the following steps:

[0092] The original infrared image is processed based on the Top-hat morphological filter, including erosion and expansion operations, and the opening and closing operations are composed of erosion and expansion operations;

[0093] If the opening operation is performed, the erosion operation is performed first and then the expansion is performed, and the formula is as follows:

[0094]

[0095] If the closed operation is performed, the expansion operation is performed first and then the corrosion is performed, and the formula is as follows:

[0096]

[0097] Among them, I is the original infrared image, S is the structured element, represents the expansion operation, Indicates a corrosion operation.

[0098]In this embodiment, the target signal in the infrared image is usually dim and mixed background clutter, which makes it difficult to detect small targets. In ord...

Embodiment 3

[0100] In this embodiment, on the basis of Embodiment 2, said step a also includes the following steps:

[0101] If the background of the original infrared image is too large and the small targets on it are regularly distributed, the background extraction can be performed based on the top-hat operation; including opening the top-hat operation and closing the top-hat operation, where:

[0102] The formula for opening the top hat operation is: OTH(x,y)=(I-S·I)(x,y);

[0103] The formula for closing the top hat operation is: CTH(x,y)=(I·S-I)(x,y).

[0104] In this embodiment, the top-hat transformation can be used to enhance the target. Opening the top-hat operation is often used to separate the patches that are brighter than the adjacent ones. When an image has a large background, and the tiny objects are more regular In some cases, the top-hat operation can be used for background extraction; closing the top-hat operation is the difference between the result image of the closed...

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

No PUM Login to View More

Abstract

The invention discloses a small target detection method based on a Markov random field and a visual comparison mechanism. The method comprises the steps of preprocessing an original infrared image ofa small target to be identified; based on a Sudoku algorithm, calculating a saliency map of the preprocessed infrared image; modeling the inter-frame difference image based on a Markov random field, and segmenting a motion area and a background area; and fusing the spatial domain saliency map and the time domain, and eliminating isolated noise points mistakenly considered as targets in the image.On the basis of a human visual contrast mechanism and a Markov random field, the moving target is detected by combining the characteristics of the time domain and the space domain, the detection of the weak target is realized, and the detection performance and effect are improved.

Description

technical field [0001] The invention relates to the technical field of computer vision, in particular to a small target detection method based on a Markov random field and a visual contrast mechanism. Background technique [0002] Infrared small target detection is an important technology in target tracking, and has always been a research hotspot in the field of infrared recognition. However, due to the relatively long distance of tracking and detection, the imaging area of ​​infrared targets is too small, and the shape and structural features are not obvious. The target is almost submerged in the The complex background makes the detection difficult to achieve. [0003] At present, infrared target detection methods can be divided into single-frame detection and multi-frame detection. The methods of multi-frame detection mainly include frame difference method, mixed Gaussian model, optical flow estimation method, etc. The multi-frame detection method requires the target bac...

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
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06K9/32G06K9/40G06T5/30G06T5/50G06T7/11G06T7/136G06T7/194
Inventor 贾海涛王颖周兰兰赵行伟
Owner 绵阳慧视光电技术有限责任公司
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com