Unlock instant, AI-driven research and patent intelligence for your innovation.

A Method of Infrared Weak and Small Target Detection

A technology of weak and small targets and detection methods, applied in the field of image processing, can solve problems such as less data and low signal-to-noise ratio, and achieve the effect of improving detection speed

Active Publication Date: 2022-03-29
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] At present, infrared weak target images are mostly used in the military field, and there is a problem of relatively little data
Moreover, due to the limitation of infrared detectors, the weak target area of ​​infrared imaging is often less than 10*10 pixels, and the signal-to-noise ratio is low

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
  • A Method of Infrared Weak and Small Target Detection
  • A Method of Infrared Weak and Small Target Detection
  • A Method of Infrared Weak and Small Target Detection

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0032] Such as figure 1 As shown, this example includes the following steps:

[0033] Step 1: training data set construction, such as figure 2 As shown; Random sampling is used to collect 8*8 background blocks in the background part of the infrared image, and when the collected background blocks overlap with the target, they are discarded. Fill the obtained background block, fill in the full coverage with a step size of 2, and obtain a 32*32 background image; use the MNIST handwritten data set as the foreground target, first change the data in the MNIST data into a black background , the image G whose number is white, the transformation formula is as follows:

[0034] G=255-I

[0035] Where I is the original MNIST handwritten dataset image, and the corresponding position of the MNIST image on the background image is replaced. The corresponding position means that MINST adds 2 to the original coordinates, so as to ensure that the target is located in the center. The above-...

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 belongs to the technical field of image processing, and in particular relates to a method for detecting small and weak infrared targets. The present invention provides a method for detecting weak and small aerial infrared targets using a convolutional neural network, using the MNIST handwritten data set to generate the required data set, aiming at the problem that the filter based on a single-frame infrared image is difficult to design, and allowing the convolutional neural network to learn the background Distribution mode, the background of the infrared image can be well suppressed through the learned filter; the step of designing the filter by oneself is avoided, because only a simple filter is used for background suppression, so the detection speed There is also a nice boost. The invention can detect images of weak and small infrared targets very well, and compares the ROC curve with other algorithms to have a relatively good effect.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a method for detecting small and weak infrared targets. Background technique [0002] Infrared imaging detection technology refers to the use of photoelectric detection technology to convert the thermal radiation signal of an object into images and graphics that can be distinguished by human vision, that is, to use the infrared radiation energy of the target to obtain relevant information of the target. The development of this technology has extended the visual range of the human eye to the infrared radiation region invisible to the naked eye. Infrared detection technology has many advantages. It adopts a continuous passive passive working method, has strong concealment and anti-interference ability, and the infrared detection system has a simple and reliable structure, small size, light weight, and easy to carry, so it plays a huge role in modern military d...

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 Patents(China)
IPC IPC(8): G06T7/246G06K9/62G06N3/04G06N3/08G06T7/223G06V10/84
CPCG06T7/246G06T7/223G06N3/08G06T2207/10048G06T2207/20024G06N3/045G06F18/29
Inventor 张萍张灵怡杨晓迪李俊宏刘西宁蒲恬
Owner UNIV OF ELECTRONICS SCI & TECH OF CHINA