Infrared imaging iced lake detection method based on MobileNetV2

A technology of infrared imaging and detection method, which is applied in the field of target detection in remote sensing and infrared image processing, and can solve problems such as low robustness and poor versatility in various scenarios

Active Publication Date: 2019-09-06
UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF7 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] In view of the problems of the above research, the purpose of the present invention is to provide a mobilenetv2-based infrared imaging icy lake detection method to solve the problem of the poor versatility and robustness of the existing detection methods based on artificially designed features for various scenarios. low problem

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
  • Infrared imaging iced lake detection method based on MobileNetV2
  • Infrared imaging iced lake detection method based on MobileNetV2
  • Infrared imaging iced lake detection method based on MobileNetV2

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0108] Step 1: Obtain a to-be-processed infrared image f of size m×n 0 (x, y) for preprocessing, where f 0 (x,y)∈R m×n , As in figure 2 Shown

[0109] Step 2: The saliency detection algorithm based on global contrast processes the preprocessed image to obtain the saliency image f 2 (x, y), where f 2 (x,y)∈R m×n , As in image 3 Shown

[0110] Step 3: Convert the saliency image f 2 (x, y) are processed to get the candidate area g i (x, y), Specific as Figure 4 Shown

[0111] Step 4: The candidate area g i (x, y) Input the fine-tuned trained MobileNetV2 convolutional neural network one by one, and get the output result y i ∈R, according to the output result y i ∈R judge each candidate area g i (x, y) Whether it is a frozen lake, for example Figure 5 Shown.

[0112] In summary, after acquiring infrared images, the present invention first preprocesses the image to eliminate noise; then uses the saliency detection algorithm based on global contrast to obtain the saliency image; then com...

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 an infrared imaging icing lake detection method based on MobileNetV2, belongs to the field of target detection in remote sensing and infrared image processing, and solves the problems that an existing detection method based on artificial design characteristics is poor in universality and low in robustness on various scenes. The method comprises the steps of obtaining a to-be-processed infrared image f0(x,y) with the size of m*n for preprocessing; processing the preprocessed image based on a global contrast saliency detection algorithm to obtain a saliency image f2(x,y);processing the saliency image f2(x,y) to obtain a candidate region gi(x,y); inputting the candidate regions gi(x,y) into the MobileNetV2 convolutional neural network subjected to fine tuning trainingone by one to obtain an output result yi belongs to R, and judging whether each candidate region gi(x,y) is an iced lake or not according to the output result yi belongs to R; the method is used fordetecting the iced lake through infrared imaging.

Description

Technical field [0001] An infrared imaging frozen lake detection method based on MobileNetV2, used for infrared imaging frozen lake detection, belongs to the target detection field in remote sensing and infrared image processing. Background technique [0002] Infrared imaging technology can realize target detection and recognition, has non-contact, strong ability to capture details, can realize long-distance detection and is not affected by obstacles such as smoke and fog, and can also realize continuous detection day and night. Therefore, in the military , Civil and other fields are widely used, especially the military value is extremely high. Among them, infrared target detection technology mainly uses the infrared radiation difference between the target and the background for target detection and recognition. In military early warning systems, infrared images generally have low contrast and blurry edges, and the presence of various false alarm sources (icing lakes, high-altit...

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 & AuthorityApplications(China)
IPC IPC(8): G06T7/00G06T7/136G06T7/187
CPCG06T7/0002G06T7/136G06T7/187G06T2207/10048G06T2207/20081G06T2207/20084G06T2207/20032G06T2207/30181Y02A90/10
Inventor宋立彭真明张鹏飞吕昱霄曹思颖刘雨菡赵学功彭凌冰何艳敏蒲恬杨春平
OwnerUNIV OF ELECTRONICS SCI & TECH OF CHINA