CN111931658A - Cooking fume identification method - Google Patents

Cooking fume identification method Download PDF

Info

Publication number
CN111931658A
CN111931658A CN202010802072.3A CN202010802072A CN111931658A CN 111931658 A CN111931658 A CN 111931658A CN 202010802072 A CN202010802072 A CN 202010802072A CN 111931658 A CN111931658 A CN 111931658A
Authority
CN
China
Prior art keywords
image
oil smoke
gray
dark channel
dark
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010802072.3A
Other languages
Chinese (zh)
Inventor
王燕
王斌龙
卢江茂
钱律求
刘文庆
李陈
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hefei Ruinatong Software Technology Development Co ltd
Original Assignee
Hefei Ruinatong Software Technology Development Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hefei Ruinatong Software Technology Development Co ltd filed Critical Hefei Ruinatong Software Technology Development Co ltd
Priority to CN202010802072.3A priority Critical patent/CN111931658A/en
Publication of CN111931658A publication Critical patent/CN111931658A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/10—Terrestrial scenes
    • F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
    • F24—HEATING; RANGES; VENTILATING
    • F24C—DOMESTIC STOVES OR RANGES ; DETAILS OF DOMESTIC STOVES OR RANGES, OF GENERAL APPLICATION
    • F24C15/00—Details
    • F24C15/20—Removing cooking fumes
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/62—Analysis of geometric attributes of area, perimeter, diameter or volume
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/56—Extraction of image or video features relating to colour

Abstract

The invention discloses a cooking oil fume identification method, which comprises the following steps: collecting an image of a cooking area in a non-oil smoke state as a reference image, and carrying out gray level difference processing on the real-time oil smoke image and the reference image to obtain a difference image; taking the dark channel value of each pixel point in the real-time oil smoke image as the channel value of the pixel point to obtain a dark channel image; and brightening and mixing the difference image and the dark channel image to obtain a composite image, and calculating the oil smoke concentration according to the information of the composite image.

Description

Cooking fume identification method
Technical Field
The invention relates to the field of oil smoke identification, in particular to a cooking oil smoke identification method.
Background
The culinary art in-process can produce a large amount of oil fumes, exposes for a long time and can directly influence healthy in the culinary art oil fume, if can real-time detection go out the oil fume among the culinary art process, then can be more intelligent control smoke ventilator's amount of wind or smoking position, bring better experience and healthy culinary art environment for the user.
In the prior art, for the detection of cooking oil fume, an oil fume sensor and a temperature sensor are mostly adopted, or a method based on general oil fume identification is adopted to detect the oil fume, but the oil fume detection based on the oil fume sensor and the temperature sensor has lower precision and smaller detectable range, and the whole condition of the oil fume is difficult to accurately describe; the image-based general oil smoke identification method can well identify the diffusion range of oil smoke, but cannot detect the thickness of the oil smoke, and cannot accurately evaluate the spatial concentration of the oil smoke.
Disclosure of Invention
In order to solve the technical problem, the invention provides a cooking fume identification method.
In order to solve the technical problems, the invention adopts the following technical scheme:
a cooking fume identification method comprises the following steps:
the method comprises the following steps: collecting an image of a cooking area in a non-oil smoke state as a reference image, and carrying out gray level difference processing on the real-time oil smoke image and the reference image to obtain a difference image;
step two: taking the dark channel value of each pixel point in the real-time oil smoke image as the channel value of the pixel point to obtain a dark channel image;
step three: and brightening and mixing the difference image and the dark channel image to obtain a composite image, and calculating the oil smoke concentration according to the information of the composite image.
Specifically, when the area where the cooking range is located in the reference image is used as an effective detection range, and when the difference image is generated in the first step, the dark channel image is generated in the second step, and the composite image is generated in the third step, the image data of each of the reference image, the real-time lampblack image, the difference image, and the dark channel image in the effective detection range is involved in the calculation.
Specifically, when the real-time oil smoke Image and the reference Image are subjected to gray level difference processing in the step one, the real-time oil smoke Image is subjected to gray level processing to generate a real-time gray level Imagerealtime_grayPerforming gradation processing on the reference Image to generate a reference gradation Imagerefer_grayDifference Imagesmoke1Gray value of each pixel point
Figure BDA0002627763560000021
Wherein
graydiff＝Imagerealtime_gray(i，j)-Imagerefer_gray(i，j)，Imagerealtime_gray(i, j) is the gray scale value of the real-time gray scale Image at the point (i, j), Imagerefer_grayAnd (i, j) is the gray scale value of the reference gray scale image at the point (i, j).
Specifically, in the second step, the minimum channel value in the red channel, the green channel and the blue channel of each pixel point in the real-time oil smoke image is used as the dark channel value of each corresponding pixel point in the dark channel image.
Specifically, in the third step, when the difference Image and the dark channel Image are brightly mixed, the dark channel Image is ImagedarkThen, the Image is synthesizedsmoke2Gray value of each pixel point
Imagesmoke2(i，j)＝max[Imagedark(i，j)，Imagesmoke1(i，j)]。
Specifically, in step three, the Image is synthesized from the synthesized Imagesmoke2When the information of (3) calculates the concentration of oil smoke, the diffusion area of oil smoke
Figure BDA0002627763560000022
Thickness of oil smoke
Figure BDA0002627763560000023
Concentration of oil fume
Figure BDA0002627763560000024
Figure BDA0002627763560000025
Wherein countnonzeroThe number of pixels with non-zero gray values in the synthetic image is defined, width is the width of the synthetic image, and height is the height of the synthetic image; m is the number of pixels of the dark channel image having a non-zero dark channel value, pixel _ darkiThe dark channel value for the ith pixel of the dark channel image having a non-zero dark channel value.
Compared with the prior art, the invention has the beneficial technical effects that:
the invention solves the problems that only the oil smoke diffusion range can be detected and the oil smoke concentration can not be evaluated in the prior art; according to the range hood, the spatial concentration of the oil smoke is estimated through the thickness and the diffusion range of the oil smoke, and the spatial concentration of the oil smoke is quantized into a corresponding range hood control strategy, so that the oil smoke processing capacity and the smoking effect of the range hood are improved.
Drawings
FIG. 1 is a flow chart of the identification method of the present invention.
Detailed Description
A preferred embodiment of the present invention will be described in detail below with reference to the accompanying drawings.
As shown in fig. 1, a cooking fume identifying method includes the following steps:
s1: and collecting an image of the cooking area in a non-oil smoke state as a reference image, and carrying out gray difference processing on the real-time oil smoke image and the reference image to obtain a difference image.
Specifically, when the real-time oil smoke Image and the reference Image are subjected to gray level difference processing in the step one, the real-time oil smoke Image is subjected to gray level processing to generate a real-time gray level Imagerealtime_grayPerforming gradation processing on the reference Image to generate a reference gradation Imagerefer_qrayDifference Imagesmoke1Gray value of each pixel point
Figure BDA0002627763560000031
Wherein
graydiff＝Imagerealtime_gray(i，j)-Imagerefer_gray(i，j)，Imagerealtime_qray(i, j) is the gray scale value of the real-time gray scale Image at the point (i, j), Imagerefer_grayAnd (i, j) is the gray scale value of the reference gray scale image at the point (i, j).
threshold is a set threshold, and if the gray difference is within the threshold range, the gray value of the difference image at the pixel point is regarded as 0; and if the gray difference value is out of the threshold range, the gray value of the difference image at the pixel point is the gray difference value.
The reference image is free of oil smoke, if oil smoke is generated in the hearth area, the gray value of a partial area of the real-time image can be changed, the two images are subjected to differential processing, and the distribution of the oil smoke can be reflected.
S2: and taking the dark channel value of each pixel point in the real-time oil smoke image as the channel value of the pixel point to obtain a dark channel image.
Specifically, in the second step, the minimum channel value in the red channel, the green channel and the blue channel of each pixel point in the real-time oil smoke image is used as the dark channel value of each corresponding pixel point in the dark channel image.
The channel is a component of the image; in the invention, each image is in an RGB mode, and the RGB mode has four channels, including a composite channel, a red channel, a green channel and a blue channel.
Dark channel prior principle: in most non-sky local areas, at least one color channel of some pixels has a very low value, in other words, the minimum value of the three channel intensities of the light in the area is a very small number.
For an arbitrary input image J, its dark channel value can be expressed by:
Figure BDA0002627763560000032
wherein JcRepresenting each channel of each pixel of the image, r, g, b representing the red, green and blue channels, respectivelyA channel; dark channel prior principles indicate that:
Jdark→0
for the image with oil smoke, the dark channel value of the image will change and will not be close to 0 any more, and in the subsequent processing, the thickness of the oil smoke will be estimated by using the dark channel value, that is, the higher the dark channel value is, the more the oil smoke is superimposed in the vertical direction, that is, the larger the thickness of the oil smoke is.
S3: and brightening and mixing the difference image and the dark channel image to obtain a composite image, and calculating the oil smoke concentration according to the information of the composite image.
Specifically, in the third step, when the difference Image and the dark channel Image are brightly mixed, the dark channel Image is ImagedarkThen, the Image is synthesizedsmoke2Gray value of each pixel point
Imagesmoke2(i，j)＝max[Imagedark(i，j)，Imagesmoke1(i，j)]。
Specifically, in step three, the Image is synthesized from the synthesized Imagesmoke2When the information of (3) calculates the concentration of oil smoke, the diffusion area of oil smoke
Figure BDA0002627763560000041
Thickness of oil smoke
Figure BDA0002627763560000042
Concentration of oil fume
Figure BDA0002627763560000043
Figure BDA0002627763560000044
Wherein countnonzeroThe number of pixels with non-zero gray values in the composite image; m is the number of pixels of the dark channel image having a non-zero dark channel value, pixel _ darkiThe dark channel value for the ith pixel of the dark channel image having a non-zero dark channel value.
The oil smoke concentration is used for representing the total amount of oil smoke in a specific three-dimensional space; the oil smoke is distributed in a three-dimensional space, but the image is two-dimensional, and the concentration of the oil smoke, namely s × depth, needs to be represented by the thickness of the oil smoke and the diffusion area of the oil smoke; and finally, carrying out normalization processing on the concentration of the oil smoke by using a sigmoid function, and mapping the concentration of the oil smoke between 0 and 1, so that the oil smoke is convenient to quantify.
Specifically, when the area where the cooking range is located in the reference image is used as an effective detection range, and when the difference image is generated in the first step, the dark channel image is generated in the second step, and the composite image is generated in the third step, the image data of each of the reference image, the real-time lampblack image, the difference image, and the dark channel image in the effective detection range is involved in the calculation.
The cooking area is complex, other interference areas are included besides the cooking bench, and if data information of the areas in the image is involved in calculation, the calculation amount is large, and meanwhile, the accuracy of the identification method is reduced.
The invention collects various images through the camera fixed on the range hood, and the images have the same size and the edges are aligned because the position and the posture of the camera are kept unchanged.
In order to reduce the calculation amount and improve the calculation accuracy, after the reference image is obtained, the position of the cooking bench is marked, and the top view of the cooking bench is generally rectangular, so that the range of the cooking bench in the reference image is taken as an effective detection range, and the coordinate of the upper left corner of the effective detection range is A (x)1，y1) The coordinate of the lower right corner is B (x)2，y2) Since the edges of the other images are aligned with the reference image, the effective detection range of each image can be obtained by using the coordinates of the point a and the point B in the other images.
In order to reduce the calculation amount and improve the identification accuracy, the data of each image in the effective detection range participates in various calculations, and the data which is not in the effective monitoring range does not participate in the calculations; for example, when calculating the gray value of each pixel point of the difference image and the gray value of each pixel point of the composite image, i ∈ [ x ]1，x2]，j∈[y1，y2](ii) a When the oil smoke concentration is calculated, width is x2-x1+1；height＝y2-y1+1。
After the oil smoke concentration is obtained through calculation, the oil smoke concentration is corresponding to the air quantity of a smoke machine, for example, when the oil smoke concentration is greater than 0.7, the maximum air suction quantity is adopted; when the concentration of the oil smoke is between 0.5 and 0.7, medium air suction quantity is adopted; when the concentration of the oil smoke is less than 0.5, the minimum air quantity is adopted; and carrying out differentiation control according to different oil smoke concentrations.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein, and any reference signs in the claims are not intended to be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may contain only a single embodiment, and such description is for clarity only, and those skilled in the art should integrate the description, and the embodiments may be combined as appropriate to form other embodiments understood by those skilled in the art.

Claims (6)

1. A cooking fume identification method comprises the following steps:
the method comprises the following steps: collecting an image of a cooking area in a non-oil smoke state as a reference image, and carrying out gray level difference processing on the real-time oil smoke image and the reference image to obtain a difference image;
step two: taking the dark channel value of each pixel point in the real-time oil smoke image as the channel value of the pixel point to obtain a dark channel image;
step three: and brightening and mixing the difference image and the dark channel image to obtain a composite image, and calculating the oil smoke concentration according to the information of the composite image.
2. The cooking fumes recognition method according to claim 1, characterized in that: and when the area where the cooking range is located in the reference image is used as an effective detection range, when a difference image is generated in the first step, a dark channel image is generated in the second step and a composite image is generated in the third step, the reference image, the real-time oil smoke image, the difference image and the dark channel image are respectively calculated from image data in the effective detection range.
3. The cooking fumes recognition method according to claim 1, characterized in that: when the real-time oil smoke image and the reference image are subjected to gray level difference processing in the first step, the real-time oil smoke image is subjected to gray level processing to generate a real-time gray level imagerealtime_grayPerforming gradation processing on the reference Image to generate a reference gradation Imagerefer_grayDifference Imagesmoke1Gray value of each pixel point
Figure FDA0002627763550000011
Wherein
graydiff＝magerealtime_gray(i,j)-Imagerefer_gray(i,j)，Imagerealtime_gray(i, j) is the gray scale value of the real-time gray scale Image at the point (i, j), Imagerefer_grayAnd (i, j) is the gray scale value of the reference gray scale image at the point (i, j).
4. The cooking fumes recognition method according to claim 1, characterized in that: and in the second step, the minimum channel value in the red channel, the green channel and the blue channel of each pixel point in the real-time oil smoke image is used as the dark channel value of each corresponding pixel point in the dark channel image.
5. The cooking fumes recognition method according to claim 1, characterized in that: in the third step, when the difference Image and the dark channel Image are brightly mixed, the dark channel Image is ImagedarkThen, the Image is synthesizedsmoke2Gray value of each pixel point
Imagesmoke2(i,j)＝max[Imagedark(i,j),Imagesmoke1(i,j)]。
6. The cooking fumes recognition method according to claim 1, characterized in that: in step three, the Image is synthesizedsmoke2When the information of (3) calculates the concentration of oil smoke, the diffusion area of oil smoke
Figure FDA0002627763550000021
Thickness of oil smoke
Figure FDA0002627763550000022
Concentration of oil fume
Figure FDA0002627763550000023
Wherein countnonzeroThe number of pixels with non-zero gray values in the synthetic image is defined, width is the width of the synthetic image, and height is the height of the synthetic image; m is the number of pixels of the dark channel image having a non-zero dark channel value, pixel _ darkiThe dark channel value for the ith pixel of the dark channel image having a non-zero dark channel value.
CN202010802072.3A 2020-08-11 2020-08-11 Cooking fume identification method Pending CN111931658A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010802072.3A CN111931658A (en) 2020-08-11 2020-08-11 Cooking fume identification method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010802072.3A CN111931658A (en) 2020-08-11 2020-08-11 Cooking fume identification method

Publications (1)

Publication Number Publication Date
CN111931658A true CN111931658A (en) 2020-11-13

Family

ID=73311550

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010802072.3A Pending CN111931658A (en) 2020-08-11 2020-08-11 Cooking fume identification method

Country Status (1)

Country Link
CN (1) CN111931658A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113063170A (en) * 2021-05-12 2021-07-02 佛山市顺德区美的洗涤电器制造有限公司 Method for identifying oil smoke, processor and range hood

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113063170A (en) * 2021-05-12 2021-07-02 佛山市顺德区美的洗涤电器制造有限公司 Method for identifying oil smoke, processor and range hood
CN113063170B (en) * 2021-05-12 2023-06-23 佛山市顺德区美的洗涤电器制造有限公司 Method for identifying lampblack, processor and range hood

Similar Documents

Publication Publication Date Title
CN106127702B (en) A kind of image defogging method based on deep learning
CN104867159B (en) A kind of digital camera sensor dust detection and stage division and device
CN105453153B (en) Traffic lights detects
Ganesan et al. Segmentation and edge detection of color images using CIELAB color space and edge detectors
Beigpour et al. Multi-illuminant estimation with conditional random fields
CN103971126B (en) A kind of traffic sign recognition method and device
JP5764238B2 (en) Steel pipe internal corrosion analysis apparatus and steel pipe internal corrosion analysis method
CN103914699A (en) Automatic lip gloss image enhancement method based on color space
CN101751672A (en) Image processing system
CN109559310A (en) Power transmission and transformation inspection image quality evaluating method and system based on conspicuousness detection
KR20020070788A (en) Region Segmentation of Color Image
JP2020161129A (en) System and method for scoring color candidate poses against color image in vision system
US10620417B2 (en) Method for generating a reflection-reduced contrast image and corresponding device
US20130170756A1 (en) Edge detection apparatus, program and method for edge detection
CN111928307B (en) Range hood with telescopic mechanism and control method thereof
CN116563799B (en) Dust online monitoring method based on video monitoring
CN107146258B (en) Image salient region detection method
CN106970709B (en) 3D interaction method and device based on holographic imaging
CN111931658A (en) Cooking fume identification method
JP2011039596A (en) Device and method for tracking image, and computer program
CN109583330B (en) Pore detection method for face photo
CN106415596B (en) image conversion based on segmentation
CN107871115A (en) A kind of recognition methods of the submarine hydrothermal solution spout based on image
CN111985436A (en) Workshop ground mark line identification fitting method based on LSD
CN110501339B (en) Cloth cover positioning method in complex environment

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination