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

Deep learning smoke recognition method for negative sample mining

A deep learning and recognition method technology, applied in character and pattern recognition, instruments, computer parts, etc., can solve the problems of poor performance of smoke recognition models, high false alarm rate, and difficulty in ensuring the accuracy of smoke detection methods. Solve the effect of insufficient samples, high accuracy, and small calculation amount

Active Publication Date: 2020-04-10
WUHAN TEXTILE UNIV
View PDF6 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The technical problem of the present invention is that the existing smoke recognition model using neural network has poor performance, high false alarm rate, and relies on a large number of training samples. The accuracy of the smoke detection method is difficult to guarantee

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
  • Deep learning smoke recognition method for negative sample mining
  • Deep learning smoke recognition method for negative sample mining
  • Deep learning smoke recognition method for negative sample mining

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0027] Such as figure 1 As shown, a deep learning smoke recognition method for negative sample mining includes the following steps,

[0028] Step 1: Collect the smoke scene image set, and extract 20 smoke templates from the smoke scene image set, including thin smoke with a small bottom and a large top, and thick smoke with extremely low transparency;

[0029] Step 2: Using the improved Faster R-CNN deep neural network, establish a smoke detection neural network model and train it, perform smoke detection and classification on the scene data set, and use the smoke-free images detected and classified as smoke as negative samples;

[0030] Step 3: Fuse the smoke template into the area that can produce smoke in the negative sample image, and generate a smoke data set for negative sample mining;

[0031] Step 4: Merge the negative sample mined smoke dataset with the smoke scene image set to form a smoke dataset;

[0032] Step 5: Use the smoke dataset to train the smoke detection...

Embodiment 2

[0038] Such as figure 1 As shown, a deep learning smoke recognition method for negative sample mining includes the following steps,

[0039] Step 1: Collect the smoke scene image set, and extract 20 smoke templates from the smoke scene image set, including thin smoke with a small bottom and a large top, and thick smoke with extremely low transparency;

[0040] Step 2: Using the Faster R-CNN deep neural network, establish and train a smoke detection neural network model, perform smoke detection and classification on the scene data set, and use the smoke-free images detected and classified as smoke as negative samples;

[0041] Step 3: Fuse the smoke template into the area that can produce smoke in the negative sample image, and generate a smoke data set for negative sample mining;

[0042] Step 4: Merge the negative sample mined smoke dataset with the smoke scene image set to form a smoke dataset;

[0043] Step 5: Use the smoke dataset to train the smoke detection neural netw...

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 deep learning smoke recognition method for negative sample mining, and the method comprises the steps: collecting a smoke scene image set, and extracting a plurality of smoketemplates from the smoke scene image set; taking a smog-free image which is detected and classified as smog as a negative sample; fusing the smoke template into an area capable of generating smoke inthe negative sample image to generate a smoke data set mined by the negative sample; combining the smoke data set mined by the negative sample with the smoke scene image set to form a smoke data set;and training a smoke detection neural network model by adopting the smoke data set, and using the smoke detection neural network model obtained by training to perform smoke detection on the scene image. The smoke identification method is high in accuracy and good in robustness; a small number of samples are adopted for training, the accuracy of smoke recognition can be guaranteed, and the problemthat the samples are insufficient is solved.

Description

technical field [0001] The invention belongs to the field of smoke detection, and in particular relates to a deep learning smoke recognition method for negative sample mining. Background technique [0002] The open-air burning of straw belongs to low-temperature incineration and incomplete combustion. The flue gas contains a large amount of carbon monoxide, carbon dioxide, nitrogen oxides, photochemical oxidants and suspended particles, causing air pollution and aggravating the occurrence of smog to a certain extent. [0003] With the rapid development of computer vision technology, more and more scenes can be recognized by computers, so more and more occasions start to use detection technology based on video analysis. In recent years, smoke detection methods based on video analysis have emerged. The smoke detection method disclosed in the Chinese patent "A Video Smoke Detection and Recognition Method Based on Transfer Learning" with the publication number CN109977790A uses...

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): G06K9/00G06K9/62
CPCG06V20/38G06F18/241
Inventor 姜明华马佩余锋周昌龙宋坤芳
Owner WUHAN TEXTILE UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com