CN102121906A - Smoke detection system as well as intelligent control system and intelligent control method of range hood - Google Patents

Abstract

The invention provides a smoke detection system and an intelligent control system using the same for a range hood. The intelligent control method comprises the following steps: analyzing captured images in a kitchen so as to perform smoke detection on the images and judge the smoke level; and setting the lampblack exhaust power of the range hood according to the smoke level, thus achieving the purpose of intelligent control on the range hood.

Description

Smoke detection system, lampblack absorber intelligence control system and lampblack absorber intelligence control method

[technical field]

The present invention relates to field of intelligent control, particularly about based on the smoke detection system of camera and the lampblack absorber intelligence control system of using this smoke detection system.

[background technology]

Lampblack absorber is one of indispensable electrical equipment in the family kitchen.At present, lampblack absorber has had very great development in profile, installation site etc., occurs but never have intelligentized product in the control of lampblack absorber.Disclose a kind of system that passes through to detect gas concentration control lampblack absorber in the patent of application number 200510048587.4, when ratio of combustion was abundant, the gas concentration in the kitchen was very low, this system cisco unity malfunction.

The major function of lampblack absorber is ventilation, the air suction smoke pipe that contains oil smoke, and discharges by flue when oil smoke in the kitchen is big, reaches and discharges the purpose that oil smoke improves air.Existing lampblack absorber all is by manually operated, and manual unlocking is manually turned off.Some people who cooks only just remembers when seeing or smell oil smoke will open lampblack absorber, and this moment, oil smoke has been diffused into air and has suffered.Finish after meal; the people can leave the kitchen very soon usually and have a meal to the dining room; this moment, oil smoke may also not taken away fully; if close lampblack absorber, will stay in the kitchen, like this by some oil smoke; manually the control lampblack absorber makes the people also will consider to close lampblack absorber when cooking finishing; and may also have oil smoke in the kitchen this moment, also needs lampblack absorber work a period of time, so just brought inconvenience to the people.

On the one hand, the power of existing lampblack absorber all is one grade or two grades that fixes, and can not regulate according to actual needs in addition.

This shows, be necessary to provide a kind of lampblack absorber control system of intelligence to improve the aforementioned disadvantages of prior art.

[summary of the invention]

The object of the present invention is to provide a kind of smoke detection system, can detect the size of amount of smoke based on camera.

Another object of the present invention is to provide a kind of intelligence control system that can control lampblack absorber according to the smokescope in kitchen automatically.

A further object of the present invention is to provide a kind of intelligence control method of lampblack absorber.

For reaching aforementioned purpose, a kind of smoke detection system of the present invention, it comprises:

Image acquisition units is used for images acquired;

The Smoke Detection unit, whether have smog, and output Smoke Detection signal if being used for detected image;

The amount of smoke analytic unit, when in detecting image, smog being arranged, the size of amount of smoke in the analysis image, and output amount of smoke analytic signal.

Further, described Smoke Detection unit detects smog by the method for image definition, and when image definition was lower than predetermined threshold, judging in the image had smog.

Further, described Smoke Detection unit selects several that smog and smogless image are arranged, employing obtains the smog sorter based on AdaBoost sorter or the training of svm classifier device, and the image of gathering is carried out feature extraction and imports the smog sorter, and whether output has the classification results of smog.

Further, described amount of smoke analytic unit is judged the size of amount of smoke by the smog area in the detected image.

For reaching aforementioned another purpose, a kind of lampblack absorber intelligence control system of the present invention, it comprises:

Image acquisition units is used to gather near the image of lampblack absorber;

The Smoke Detection unit, whether be used for detected image has smog;

The amount of smoke analytic unit, when in detecting image, smog being arranged, the amount of smoke size in the analysis image;

Lampblack absorber start-up control unit, the result who detects according to the Smoke Detection unit starts lampblack absorber;

The lampblack absorber power control unit is controlled the operating power of lampblack absorber according to the amount of smoke result of amount of smoke analytic unit analysis.

Further, described Smoke Detection unit detects smog by the method for image definition, and when image definition was lower than predetermined threshold, judging in the image had smog.

Further, described Smoke Detection unit selects several that kitchen range top image of oil smoke and no oil smoke is arranged, employing obtains the oil smoke sorter based on AdaBoost sorter or the training of svm classifier device, and the image of gathering is carried out feature extraction and imports the oil smoke sorter, and whether output has the classification results of oil smoke.

Further, analyze amount of smoke less than predetermined when starting threshold value when the amount of smoke analytic unit, the amount of smoke analytic unit is closed lampblack absorber by lampblack absorber start-up control unit.

Further, described lampblack absorber intelligence control system, it further comprises an alarm unit, when the amount of smoke analytic unit is analyzed amount of smoke greater than alarm threshold value, sends warning message by alarm unit.

For reaching an aforementioned purpose again, a kind of lampblack absorber intelligence control method of the present invention, it comprises:

Image acquisition units is gathered near the image the lampblack absorber;

In the Smoke Detection unit detected image whether smog and output signal are arranged;

Lampblack absorber start-up control unit is according to the startup of the output signal of Smoke Detection unit control lampblack absorber/close;

After the lampblack absorber control module starts lampblack absorber, the size of amount of smoke in the amount of smoke analytic unit analysis image;

The lampblack absorber power control unit is according to the power of the size control lampblack absorber of described amount of smoke.

The present invention carries out Smoke Detection by the image in the kitchen of analyze taking to image, and the rank of judgement smog, according to the degree of smog the smoking power of smoke exhaust ventilator is set, and reaches the purpose of Based Intelligent Control smoke exhaust ventilator.

[description of drawings]

Fig. 1 is the structured flowchart of an embodiment of smoke detection system of the present invention.

Fig. 2 is the structured flowchart of an embodiment of lampblack absorber intelligence control system of the present invention.

Fig. 3 is the workflow diagram of an embodiment of lampblack absorber intelligence control system of the present invention.

[embodiment]

Alleged herein " embodiment " or " embodiment " are meant special characteristic, structure or the characteristic that can be contained at least one implementation of the present invention.Different in this manual local " in one embodiment " that occur not are all to refer to same embodiment, neither be independent or optionally mutually exclusive with other embodiment embodiment.In addition, represent the sequence of unit in method, process flow diagram or the functional block diagram of one or more embodiment and revocablely refer to any particular order, also be not construed as limiting the invention.

See also shown in Figure 1ly, it shows the structured flowchart of the smoke detection system of one embodiment of the invention.As shown in the figure, the smoke detection system 100 of one embodiment of the invention, it comprises image acquisition units 1, Smoke Detection unit 2 and amount of smoke analytic unit 3.

Image acquisition units 1 is used to gather ambient image, for example kitchen or compartment or other environment etc.

The image that Smoke Detection unit 2 is gathered according to image acquisition units 1, image is detected, whether have smog in the detected image, when smog is arranged, the sharpness of the image that image acquisition units 1 is taken can reduce, and can judge whether to exist smog by the clear picture degree.The definition mode of given image definition and threshold value when image definition during greater than threshold value, are thought not have smog, otherwise think and have smog, then send the testing result signal that has smog in the image.The definition of sharpness R has multiple mode, for example sets I (x) and is the function of gradation of image, supposes that the gray scale that a is ordered in the image is I (a), and the gray scale that b is ordered is I (b), and the rate of gray level that a is ordered is

Wherein d is a little distance, I (a+d) is and a distance is the gray scale of the point of d, and the value of d can get 1 usually, and then a rate of gray level of ordering is I (a+1)-I (a), the rate of gray level that same b is ordered is I (b+1)-I (b), and the formula of sharpness R can be defined as:

$R=\frac{\underset{x=a}{\overset{b}{\mathrm{\Σ}}}{(I(x+1)-I\left(x\right))}^2}{|I\left(b\right)-I\left(a\right)|}$

Be appreciated that ground, for the camera that is used to take kitchen fume, its position, after, exposure, focal length etc. are determined, no longer change, so the picture material of its shooting be constant, variation just above the smog situation.Also can be when calculating sharpness R by in image, selecting several sampled points, the sharpness sum of calculating all sampled points is represented the sharpness of present image.

Except the method that adopts image definition to analyze, can also adopt the method for training smog sorter.Select several that smog and smogless image are arranged, by there being non smoke to be divided into two groups, constitute the positive and negative sample of smog sorter 21, then image is carried out feature extraction, feature should guarantee to comprise the relevant information of smog, as adopt the contrast constitutive characteristic vector of a plurality of points, perhaps adopt the acutance constitutive characteristic vector of multiple spot, perhaps sharpness proper vector etc.Then the input of the proper vector of positive and negative sample set is wanted in the training module of smog sorter, can be adopted sorter, perhaps svm classifier device etc. based on AdaBoost.After training obtains smog sorter 21, the image of taking is carried out feature extraction and imports smog sorter 21, whether output has the classification results of smog.

Smoke Detection unit 2 can only detect whether smog is arranged in the image, and amount of smoke analytic unit 3 then can further be determined the size of amount of smoke, and output amount of smoke analytic signal.

Because smog is that apparent in view shape is arranged in image, amount of smoke analytic unit 3 can obtain current smog shape by extracting, thereby determines the amount of current smog, and according to amount of smoke output amount of smoke analytic signal.

Concrete method is that at first obtaining a width of cloth does not have the image of smog visual as reference, then current frame image and reference picture is subtracted each other, and obtains the image-region of smog, analyzes this regional area then, determines current amount of smoke according to area.If area is very little, illustrate that amount of smoke is smaller, if the bigger explanation amount of smoke of area is bigger, by different threshold values is set, when the smog area reaches different threshold value, the amount of smoke analytic signal that the output of amount of smoke analytic unit is different.

See also shown in Figure 2ly, it shows that the present invention adopts the structured flowchart of an embodiment of the lampblack absorber intelligence control system of aforementioned smoke detection system.As shown in the figure, lampblack absorber intelligence control system of the present invention 200 it comprise: image acquisition units 1, Smoke Detection unit 2, amount of smoke analytic unit 3, lampblack absorber start-up control unit 4, lampblack absorber power control unit 5, alarm unit 6.

Normal kitchen of using can produce oil smoke when cooking, and the main positions of oil smoke is kitchen range top top.In order to siphon away oil smoke, the user installs smoke exhaust ventilator usually near the kitchen range top, as is installed in the Chinese style or the European smoke exhaust ventilator of kitchen range top top, is installed in the side-absorption type smoke exhaust ventilator of kitchen range top side top.Use dissimilar smoke exhaust ventilators, the movement locus of oil smoke is different, the top formula be to make progress by the kitchen range top, enter lampblack absorber, the side formula is to enter lampblack absorber by the kitchen range top on side.In order to detect oil smoke, the image acquisition units 1 of lampblack absorber intelligence control system of the present invention is used to gather the smog image in kitchen, and it mainly comprises the camera of the image of a shooting kitchen range top.The position of camera should guarantee to photograph the overall picture of the kitchen range top, and is positioned at the less place of oil smoke, as the outside, bottom of top formula lampblack absorber, the panel upper outside of side formula lampblack absorber.The position of adjustment camera is also fixed.And focal length is suitable, and phenomenons such as defocusing blurring can not appear in the kitchen range top image of shooting, because oil smoke can bring the fuzzy of image, also can bring the fuzzy of image and defocus, if two kinds of fuzzy being mixed in together are which kind of reason is difficult to distinguish one from the other be on earth, can improve the difficulty of Smoke Detection.

Lampblack absorber start-up control unit 4 be used to control lampblack absorber 7 startup, close.

The effect of Smoke Detection unit 2 is the kitchen range top treatment of picture and the analyses of taking by to camera, and whether confirm has smog on the kitchen range top.Discern the method that the method whether smog is arranged on the kitchen range top can adopt image definition to analyze.

When on the kitchen range top oil smoke being arranged, the sharpness of the image of shooting has reduced, and can judge whether to exist oil smoke by the clear picture degree.The definition mode of given image definition and threshold value when image definition during greater than threshold value, are thought not have oil smoke, otherwise think and have oil smoke, need to start smoke exhaust ventilator.The definition of sharpness has the definition of multiple mode sharpness R that multiple mode is arranged, and for example sets I (x) and is the function of gradation of image, supposes that the gray scale that a is ordered in the image is I (a), and the gray scale that b is ordered is I (b), and the rate of gray level that a is ordered is

Wherein d is a little distance, I (a+d) is and a distance is the gray scale of the point of d, and the value of d can get 1 usually, and then a rate of gray level of ordering is I (a+1)-I (a), the rate of gray level that same b is ordered is I (b+1)-I (b), and the formula of sharpness R can be defined as:

$R=\frac{\underset{x=a}{\overset{b}{\mathrm{\Σ}}}{(I(x+1)-I\left(x\right))}^2}{|I\left(b\right)-I\left(a\right)|}$

Be appreciated that ground, for the camera that is used to take kitchen fume, its position, after, exposure, focal length etc. are determined, no longer change, so the picture material of its shooting be constant, variation just above the smog situation.Also can be when calculating sharpness R by in image, selecting several sampled points, the sharpness sum of calculating all sampled points is represented the sharpness of present image.

Except the method that adopts image definition to analyze, can also adopt the method for training oil smoke sorter.Select several that kitchen range top image of oil smoke and no oil smoke is arranged, by having or not oil smoke to be divided into two groups, constitute the positive and negative sample of oil smoke sorter 21, then image is carried out feature extraction, feature should guarantee to comprise the relevant information of oil smoke, as adopt the contrast constitutive characteristic vector of a plurality of points, perhaps adopt the acutance constitutive characteristic vector of multiple spot, perhaps sharpness proper vector etc.Then the input of the proper vector of positive and negative sample set is wanted in the training module of oil smoke sorter, can be adopted sorter, perhaps svm classifier device etc. based on AdaBoost.After training obtains oil smoke sorter 21, the kitchen range top image of taking is carried out feature extraction and imports oil smoke sorter 21, whether output has the classification results of oil smoke.

Smoke Detection unit 2 can only detect whether smog is arranged in the image, and amount of smoke analytic unit 3 then can further be determined the size of amount of smoke, basis for estimation is provided for the lampblack absorber power control unit 5 of back.

After lampblack absorber starts, because the influence that the cigarette machine absorbs, though remain the kitchen range top image of oil smoke, but the kitchen range top of this image when being not activated lampblack absorber has the image of oil smoke that a great difference has been arranged, because after lampblack absorber started, the oil smoke on the kitchen range top can be attracted to smoke exhaust ventilator, therefore, oil smoke can be more concentrated.Relatively the oil smoke of Ji Zhonging is that apparent in view shape is arranged in image, can be by extract obtaining current oil smoke shape, and oil smoke is to the full level of lampblack absorber, thus determine the amount of current oil smoke, and determine the required power of bleeding according to amount of smoke.

Concrete method is that at first obtaining a width of cloth does not have the kitchen range top image of oil smoke visual as reference, then current frame image and reference picture is subtracted each other, and obtains the image-region of oil smoke, analyzes this regional area then, determines current amount of smoke according to area.If area is very little, illustrate that amount of smoke is smaller, only need smaller power just much of that, when the smog area is bigger, even exceeded kitchen range top zone, illustrate that current power is not enough, need to adopt bigger power.

Lampblack absorber power control unit 5 is controlled the power of lampblack absorber 7 according to the amount of smoke size of amount of smoke analytic unit analysis, during smoke exhaust ventilator work, the power of smoking is not high more good more, and this is because the high more power consumption more of power, and sound is big more, and big more to the degree of wear of motor.Therefore, the power of smoking is not high more good more, but it is just passable to satisfy the needs of smoking.

Here the mode that adopts function is described the relation of smoking power and smog area.

Smog area s, power demand p, then

p＝f(s)

Usually, f can adopt linear mode.

If s is less than preset threshold value a small amount of E, then thinking has not had cigarette, just closes smoke exhaust ventilator.

If the power bracket of smoke exhaust ventilator is [p _Min, p _Max], then can adopt following function mode:

$\left\{\begin{array}{ccc}\mathrm{if}& E<s\&le;s_0& p=p_{\min }\\ \mathrm{if}& s_0<s<s_1& p=k(s-s_0)+p_{\min }\\ \mathrm{if}& s\&GreaterEqual;s_1& p=p_{\max }\end{array}\right.$

If different with current power, then need power is adjusted by the power that function calculation obtains.The process of adjusting should be mild, rather than rolls up quickly or reduce power, because can add the rate of decay of speed motor like this.Can establish one and adjust step-length, when power need be adjusted, progressively adjust with this step-length.In addition, for the judgement that whether needs to adjust power, just do not need neither definitely equate to adjust, but an interval range can be set, when the difference of power demand that calculates and current power satisfies this interval range and requires, do not adjust, when exceeding this scope, adjust again.

If do not adopt the mode of continuous adjusting, the mode that also can adopt classification to regulate, promptly the smoking power of smoke exhaust ventilator is divided into several grades, selects suitable gear according to the situation of smog.This mode only needs to increase several threshold values on the basis of method in the above and just can realize.Present many lampblack absorbers are supported two grades of adjustings, but need manually regulate, and utilize this method can realize automatic adjusting.

The quantity of smog is very big in the kitchen, when surpassing conventional value, can think that the kitchen has caught fire, and perhaps pot has dryouied etc., and can report to the police this moment by alarm unit.Be s＞s _ThThe time, carry out danger warning.The warning message of alarm unit can be various warning messages such as audible alarm, light flash warning, text prompt warning.

See also shown in Figure 3ly, it shows the workflow diagram of lampblack absorber intelligence control system of the present invention.As shown in the figure, the course of work of lampblack absorber intelligence control system of the present invention is:

Step S301; Image acquisition units is gathered near the image the kitchen range top;

Step S302: image acquisition units is transferred to the Smoke Detection unit with near the image the kitchen range top that collects, in the Smoke Detection unit detected image whether smog is arranged, concrete method can be the method that adopts the method for earlier figures image sharpness analysis or adopt training oil smoke sorter.

Step S303:, then return step S301 and continue images acquired when the Smoke Detection unit detects when not having smog in the kitchen.

Step S304; When the Smoke Detection unit detects when in the kitchen smog being arranged, output signal to lampblack absorber start-up control unit, the lampblack absorber control module starts lampblack absorber.

Step S305: after the lampblack absorber control module started lampblack absorber, the amount of smoke analytic unit was analyzed the size of amount of smoke.

Step S306: the different threshold values that different amount of smoke can be set, for example start threshold value, one grade of threshold value, two grades of threshold values etc., when amount of smoke during greater than predetermined different threshold value, amount of smoke analytic unit output analysis result is given oil smoke acc power control module, by the power of lampblack absorber power control unit control lampblack absorber.

Step S307: when amount of smoke less than predetermined when starting threshold value, amount of smoke analytic unit output analysis result is given lampblack absorber start-up control unit, closes lampblack absorber by lampblack absorber start-up control unit.

Step S308: when amount of smoke surpassed alarm threshold value, amount of smoke analytic unit notice alarm unit sent alerting signal.

Above-mentioned explanation has fully disclosed the specific embodiment of the present invention.It is pointed out that and be familiar with the scope that any change that the person skilled in art does the specific embodiment of the present invention does not all break away from claims of the present invention.Correspondingly, the scope of claim of the present invention also is not limited only to previous embodiment.

Claims (10)

1. smoke detection system, it comprises:

Image acquisition units is used for images acquired;

The Smoke Detection unit, whether have smog, and output Smoke Detection signal if being used for detected image;

The amount of smoke analytic unit, when in detecting image, smog being arranged, the size of amount of smoke in the analysis image, and output amount of smoke analytic signal.

2. smoke detection system as claimed in claim 1 is characterized in that: described Smoke Detection unit detects smog by the method for image definition, and when image definition was lower than predetermined threshold, judging in the image had smog.

3. smoke detection system as claimed in claim 1, it is characterized in that: described Smoke Detection unit selects several that smog and smogless image are arranged, employing obtains the smog sorter based on AdaBoost sorter or the training of svm classifier device, the image of gathering is carried out feature extraction and imports the smog sorter, and whether output has the classification results of smog.

4. smoke detection system as claimed in claim 1 is characterized in that: described amount of smoke analytic unit is judged the size of amount of smoke by the smog area in the detected image.

5. lampblack absorber intelligence control system, it comprises:

Image acquisition units is used to gather near the image of lampblack absorber;

The Smoke Detection unit, whether be used for detected image has smog;

The amount of smoke analytic unit, when in detecting image, smog being arranged, the amount of smoke size in the analysis image;

Lampblack absorber start-up control unit, the result who detects according to the Smoke Detection unit starts lampblack absorber;

The lampblack absorber power control unit is controlled the operating power of lampblack absorber according to the amount of smoke result of amount of smoke analytic unit analysis.

6. lampblack absorber intelligence control system as claimed in claim 5 is characterized in that: described Smoke Detection unit detects smog by the method for image definition, and when image definition was lower than predetermined threshold, judging in the image had smog.

7. lampblack absorber intelligence control system as claimed in claim 5, it is characterized in that: described Smoke Detection unit selects several that kitchen range top image of oil smoke and no oil smoke is arranged, employing obtains the oil smoke sorter based on AdaBoost sorter or the training of svm classifier device, the image of gathering is carried out feature extraction and imports the oil smoke sorter, and whether output has the classification results of oil smoke.

8. lampblack absorber intelligence control system as claimed in claim 5 is analyzed amount of smoke less than predetermined when starting threshold value when the amount of smoke analytic unit, and the amount of smoke analytic unit is closed lampblack absorber by lampblack absorber start-up control unit.

9. lampblack absorber intelligence control system as claimed in claim 5, it further comprises an alarm unit, when the amount of smoke analytic unit is analyzed amount of smoke greater than alarm threshold value, sends warning message by alarm unit.

10. lampblack absorber intelligence control method, it comprises:

Image acquisition units is gathered near the image the lampblack absorber;

In the Smoke Detection unit detected image whether smog and output signal are arranged;

Lampblack absorber start-up control unit is according to the startup of the output signal of Smoke Detection unit control lampblack absorber/close;

After the lampblack absorber control module starts lampblack absorber, the size of amount of smoke in the amount of smoke analytic unit analysis image;

The lampblack absorber power control unit is according to the power of the size control lampblack absorber of described amount of smoke.

Images