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

Flame detection method and device

A flame detection and flame technology, applied in the field of digital image processing, can solve the problems of high confidence in false detection, unsatisfactory false detection rate, and false detection, and achieve the effect of reducing false detection rate, improving accuracy rate, and low demand

Active Publication Date: 2021-11-23
小视科技(江苏)股份有限公司
View PDF3 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the existing detection network generally detects a single image, and the form of flames on a single image may be very similar to various bright and red lights, which is very easy to cause false detection in actual use
In the infrared scene, the flame is almost the same as the light. The flame will flicker irregularly, but the light will not. Therefore, it is completely impossible to distinguish between the flame and the light in a single image. The confidence of false detection is extremely high, and the flame detection cannot be balanced. False detection and recall rate, making it impossible to use in actual production on a large scale
[0003] The existing video-based flame detection technology realizes flame detection in video by extracting the static and dynamic features of the video, taking into account the characteristics of the flame's spatial movement and time continuity, etc., and can perform flame detection more effectively; but its static, dynamic Feature extraction algorithms and classification models are often extremely complex, for example, complex processing methods such as multi-layer wavelet decomposition, Hu invariant moments, Kalman filter, Markov model, or neural networks with extremely complex structures are required. Not good, the demand for hardware and software resources is extremely high, and the false detection rate of actual detection is not satisfactory

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
  • Flame detection method and device
  • Flame detection method and device
  • Flame detection method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0024] Aiming at the deficiencies of the existing technology, the solution of the present invention is based on the flame detection results based on single-frame images, and according to the physical principle of flame changes, by simply processing the time series of image frames, the high-frequency beating The flame detection results are retained, and low-frequency changes such as personnel movement and background movement are eliminated, thereby greatly reducing the false detection rate. At the same time, the single-frame image flame detection technology has good real-time performance and low demand for software and hardware resources. The advantage of being easy to implement.

[0025] Specifically, the flame detection method proposed by the present invention is as follows:

[0026] Use the flame detection method based on a single frame image to perform preliminary frame-by-frame detection of the video. When the flame is initially detected, the gray-scale images of the flame...

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 flame detection method. The flame detection method based on a single frame image is used to perform preliminary frame-by-frame detection of a video. Extract the grayscale image of the flame area respectively, and obtain the frame difference image of the grayscale image of the flame area between adjacent frame images; weight the superimposed image of all frame difference images and the reverse image of all frame difference images and, the flame detection probability map is obtained. If the number of pixels in the flame detection probability map whose pixel values ​​exceed the preset threshold exceeds the preset number, it is determined that the preliminary detection result is correct, otherwise it is a false detection. The invention also discloses a flame detection device. Compared with the prior art, the present invention not only has higher detection accuracy, but also has better real-time performance, and has lower requirements for software and hardware resources.

Description

technical field [0001] The invention relates to digital image processing technology, in particular to a flame detection method. Background technique [0002] With the popularization of video surveillance equipment and the development of digital image processing technology, the flame detection technology based on visual image has also been developed rapidly. Due to the weak texture, changeable shape, and diversity of flame images, traditional flame detection algorithms mainly identify flames through color and brightness, plus logic for multiple confirmations. In recent years, due to the sudden emergence of convolutional neural networks, flame detection can use detection networks to detect flames, which can greatly improve detection accuracy and efficiency. However, the existing detection network generally detects a single image, and the form of flames on a single image may be very similar to various bright and red lights, which is very easy to cause false detection in actual...

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): G06T7/00G06N3/08
CPCG06N3/08G06T7/0002G06T2207/10016G06T2207/10048G06T2207/20081G06T2207/20084G06T2207/30232
Inventor 杨帆白立群胡建国
Owner 小视科技(江苏)股份有限公司
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