WO2015117427A1

WO2015117427A1 - Dirt detection method, device and terminal

Info

Publication number: WO2015117427A1
Application number: PCT/CN2014/090111
Authority: WO
Inventors: 曹争光
Original assignee: 中兴通讯股份有限公司
Priority date: 2014-08-25
Filing date: 2014-10-31
Publication date: 2015-08-13
Also published as: CN105389577A

Abstract

Provided are a dirt detection method, device and terminal. The method comprises: a scanning step, conducting scanning imaging on a medium to be analysed according to requirements of a pre-set frame number; a recognition step, according to the distribution of pixels of various colours of an imaging result of a current frame number, recognizing background colours of the imaging result of the current frame number; an analysis step, according to the distribution of pixels of non-background colours of the imaging result of the current frame number, determining whether there is dirt in the imaging result of the current frame number; a photographing step, if it is determined that there is dirt in the imaging result of the current frame number, photographing the imaging result of the current frame number to obtain photographing results; and an evaluation step, after the scanning step is finished, according to all the obtained photographing results, evaluating the adhesion condition of the dirt of the medium to be analysed.

Description

Method, device and terminal for detecting dirt

Technical field

The invention relates to the field of graphic recognition, in particular to a method, a device and a terminal for detecting dirt.

Background technique

With the improvement of the quality of life, people's requirements for sanitary conditions are getting higher and higher. Especially the health of children. Children's perception of the surrounding environment is mainly touch, and inevitably, they will be exposed to a lot of dirt. Parents need a simple operation device to confirm the sanitary condition of children's contacts at any time. At present, common terminal tools such as mobile phones have high-pixel cameras, which can clearly image the object to be magnified, and combined with the current computer graphics recognition technology, can identify dirt that is difficult to be recognized by the naked eye.

Therefore, how to implement the dirt detection on the commonly used terminal equipment is a technical problem that needs to be solved at present.

Summary of the invention

In order to solve the prior art problems, the present invention is intended to provide a method, apparatus, and terminal for detecting dirt.

Embodiments of the present invention provide a method for detecting dirt, including:

Scanning step, scanning and imaging the medium to be tested according to the preset number of frames;

The identifying step identifies the background color of the imaging result of the current frame number according to the pixel distribution of each color of the imaging result of the current frame number;

The analyzing step determines whether the imaging result of the current frame number has dirt according to the pixel distribution of the non-background color of the imaging result of the current frame number;

The photographing step, if it is determined that the imaging result of the current frame number has dirt, the current frame number is Taking the imaging result and taking a picture to obtain a photograph result;

The evaluating step, after the end of the scanning step, evaluates the adhesion of the dirt of the medium to be tested based on all the obtained photographing results.

In addition, another embodiment of the present invention provides a device for detecting dirt, including:

The scanning module is configured to scan and image the medium to be tested according to a preset number of frames;

An identification module configured to identify a background color of an imaging result of the current frame number according to a pixel distribution of each color of the imaging result of the current frame number;

The analyzing module is configured to determine, according to the pixel distribution of the non-background color of the imaging result of the current frame number, whether the imaging result of the current frame number has dirt;

The photographing module is configured to: if it is determined that the imaging result of the current frame number has dirt, photograph the imaging result of the current frame number to obtain a photographing result;

The evaluation module is configured to evaluate the adhesion of the dirt of the medium to be tested according to all the obtained photographing results after the scanning step ends.

Further, another embodiment of the present invention provides a terminal including the above-described apparatus for detecting dirt.

The beneficial effects of the above technical solution of the present invention are as follows:

The solution of the embodiment of the invention can detect and analyze the dirt of the target object, in particular, the application scenario for determining the sanitary condition. In addition, the entire solution can be executed on terminals that are currently available on the market, so the cost is low and the penetration rate is high.

DRAWINGS

1 is a schematic diagram of steps of a method for detecting dirt according to an embodiment of the present invention;

Figure 2 is a flow chart showing a method of detecting the present invention;

FIG. 3 is a schematic structural diagram of a device for detecting dirt according to an embodiment of the present invention.

detailed description

The technical problems, the technical solutions, and the advantages of the present invention will be more clearly described in the following description.

As shown in FIG. 1 , an embodiment of the present invention provides a method for detecting a dirt, including:

Scanning step 11, performing scanning imaging on the medium to be tested according to the preset number of frames;

The identifying step 12 is to identify the background color of the imaging result of the current frame number according to the pixel distribution of each color of the imaging result of the current frame number;

In the analyzing step 13, determining, according to the pixel distribution of the non-background color of the imaging result of the current frame number, whether the imaging result of the current frame number has dirt;

In the photographing step 14, if it is determined that there is dirt in the imaging result of the current frame number, the imaging result of the current frame number is photographed, and the photographing result is obtained;

After entering the imaging result of the next frame, the recognition step 12 to the photographing step 14 are repeatedly performed. Until the scanning step 11 ends.

In the evaluation step 15, after the scanning step 11 is completed, the adhesion of the dirt of the medium to be tested is evaluated based on all the obtained photographing results.

It can be known from the above description that the method of the embodiment of the invention can detect and analyze the dirt of the target object, in particular, the application scenario for determining the sanitary condition. In addition, the entire method can be completely executed on terminals currently on the market, so the cost is low and the penetration rate is high.

Specifically, the foregoing identifying step 12 includes:

Step 121: Determine a pixel ratio of each color; wherein, a pixel ratio of a certain color=the number of consecutive pixels of the color/the number of all pixels in the imaging result of the current frame number;

Step 122: Determine a color whose pixel ratio is greater than a first preset threshold as a background color of an imaging result of the current frame number.

In the case where the dirt coverage area is large, there is no need to worry that the above steps 121 and 122 may incorrectly determine the color of the stain as the background color. Because in this scenario, the user is completely visible to the naked eye. Visually check the attachment of dirt.

In addition, in the above analysis step 13, the specific includes:

Step 131, determining a sum of pixel ratios of non-background colors;

Step 132: When the sum of the pixel ratios of the non-background colors is greater than the second preset threshold, determining that the imaging result of the current number of frames has dirt.

In the evaluating step 15, the method specifically includes:

Step 151: Determine a dirt adhesion degree of the medium to be tested according to the photographing result; the dirt adhesion degree=the number of pixels of the non-background color of all the photographing results/the number of all the pixels of all the photographing results;

Step 152, evaluating the adhesion of the dirt of the medium to be tested according to the degree of soil adhesion.

The above steps 151 and 152 are evaluations of the adhesion of the unsuitable dirt. Therefore, the area of the dirt is much smaller than the total area of the entire photographing result. The evaluation criteria can be referred to Table 1:

X值X value	污物附着度Soil adhesion
X值X value	污物附着度Soil adhesion	x≤5％X≤5%	无污染No pollution
6％<x≤10％6% < x ≤ 10%	轻微污染 Slight pollution	x≤5％X≤5%	无污染No pollution
6％<x≤10％6% < x ≤ 10%	轻微污染 Slight pollution	11％<x≤15％11% < x ≤ 15%	一般污染General pollution
16<x≤20％16<x≤20%	中度污染Moderately polluted	11％<x≤15％11% < x ≤ 15%	一般污染General pollution
16<x≤20％16<x≤20%	中度污染Moderately polluted	20％<x20%<x	重度污染Severe pollution

Table I

In Table 1, x is the degree of soil adhesion, and the pixel point is regarded as the area, X = (the area of the dirt in the photographing result 1 + the area of the dirt in the photographing result 2 + the area of the stain in the photographing result 3 +). .......+Photographing result N the area of dirt)/(Photographing result 1 total area + total area of photographing result 2 + total area of photographing result 3 +........ + taking photo The total area of N results). That is, the density of the average stain of the plurality of photographing results was taken to evaluate the degree of adhesion.

In addition, like a towel waiting to measure the medium, may have a pattern or pattern, in order to avoid these The graphic is mistaken for dirt, and the identifying step 12 further includes:

Step 123: Configure a coordinate system for all pixel points of the imaging result of the current frame number;

Step 124: Determine a graphic of each color according to coordinates corresponding to each color;

Step 125: Determine a color corresponding to the graphic that meets the preset rule as a background color;

The preset rule includes: a circular (corresponding to a speckle pattern) and/or a linear shape (corresponding to a stripe pattern),

Of course, the above-mentioned preset rules may also include common rule patterns such as diamonds and squares, which are not listed here.

The method of the present invention will be specifically described below in conjunction with an implementation.

In this implementation mode, the execution subject is a mobile phone, and the white towel is detected as an example. The specific flow diagram is shown in FIG. 2:

Step 21: The user expands a white towel to determine the scanned area, and aligns the camera with the scanned object;

Step 22: Start the camera of the terminal and the scanning program, and enter the scanning interface.

Step 23: At this time, the program corresponding to the analysis step is started, and the white color is recognized as the background color. After the user confirms the background color, the handheld terminal moves and scans, and the camera scans slowly in the scan area determined by the user; if the camera preview screen is detected There is a foreign matter having a large difference from the background color. First, it is analyzed whether the foreign matter is a pattern or a pattern on the towel. If not, the dirt is found to be found, and step 24 is performed. If yes, it is determined that the current area is found in other areas. Continue to step 22;

Step 24: Release a photo instruction to shoot it;

Step 25: Save the captured photos to the terminal data storage module.

Step 26: After the user completes the scanning, manually close the scanning interface and enter the result output interface. If there is no dirt, the display is “clean”, and if there is dirt, the captured dirt photo is displayed;

Step 27: performing an evaluation step of determining a dirt adhesion degree of the white towel according to the photographing result, and evaluating the adhesion of the dirt of the medium to be tested according to the dirt adhesion degree;

Step 28: Output the stain adhesion determined by the evaluation step.

In addition, as shown in FIG. 3, a device for detecting dirt according to an embodiment of the present invention includes:

The scanning module 31 is configured to scan and image the medium to be tested according to a preset number of frames;

The identifying module 32 is configured to identify a background color of the imaging result of the current frame number according to a pixel distribution of each color of the imaging result of the current frame number;

The analyzing module 33 is configured to determine, according to the pixel distribution of the non-background color of the imaging result of the current frame number, whether the imaging result of the current frame number has dirt;

The photographing module 34 is configured to: if it is determined that there is dirt in the imaging result of the current frame number, take a photo of the imaging result of the current frame number to obtain a photographing result;

The evaluation module 35 is configured to evaluate the adhesion of the dirt of the medium to be tested according to all the obtained photographing results after the scanning step ends.

As can be seen from the above description, the device of the present invention is capable of detecting and analyzing the dirt of the target object, particularly in a hygienic detection scenario. In addition, the terminals commonly used in the market are basically equipped with the above modules, so the cost is low and the penetration rate is high.

Specifically, on the basis of the above, the identification module 32 includes:

a first determining sub-module configured to determine a pixel ratio of each color; wherein, a pixel ratio of a certain color=the number of consecutive pixels of the color/the number of all pixels in the imaging result of the current frame number;

And a second determining submodule configured to determine a color whose pixel ratio is greater than the first preset threshold as the background color of the imaging result of the current frame number.

Specifically, based on the foregoing, the analyzing module 33 includes:

a third determining submodule configured to determine a sum of pixel proportions of non-background colors;

a fourth determining submodule configured to: when a sum of pixel ratios of non-background colors is greater than a second preset threshold When the value is determined, it is determined that the imaging result of the current frame number has a stain.

Specifically, based on the above, the evaluation module 35 includes:

Processing the sub-module, determining, according to the photographing result, the dirt adhesion degree of the medium to be tested; the dirt adhesion degree=the number of pixels of the non-background color of all the photographing results/the number of all the pixels of all the photographing results;

The evaluation sub-module evaluates the adhesion of the dirt of the medium to be tested according to the degree of soil adhesion.

Specifically, the identification module 32 further includes:

a fifth determining submodule configured to configure a coordinate system for all pixels of the imaging result of the current frame number;

a sixth determining submodule configured to determine a graphic of each color according to coordinates corresponding to each color;

a seventh determining submodule configured to determine a color corresponding to the graphic that meets the preset rule as a background color;

The preset rule includes: a circle and/or a line shape.

In addition, another embodiment of the present invention further provides a terminal comprising: the above-mentioned dirt detecting device. The terminal may be a device such as a PAD with a camera or a mobile phone.

The scanning module 31 includes a high-definition pixel macro camera, which is mainly configured to shoot at a small distance. Since the scanned object is generally not too large in area, the macro camera can obtain a relatively clear enlarged image at a small distance, which can Guaranteed to be accurate and fast. During the scanning process, the macro camera maintains the preview state. Because it is in the macro mode, it does not need to adjust the focal length, and the anti-shake system can ensure that a relatively clear photo is obtained after receiving the camera instruction, and the photograph taken is recorded. Coordinate information.

The identification module 32 and the analysis module 33 are respectively a graphic recognition system and a graphic analysis system in the terminal. The function of the graphic recognition system is to give the scanning module 31 a photographing instruction when the scanning module After starting work at 31, the graphics recognition system started working.

The graphic recognition system marks consecutive pixels of similar colors in the scanning area. When the number of marked pixels exceeds one third of the number of pixels included in one frame of image, the color range is used as the image of the frame. The background color, if a pixel point set having a large difference from the background color is found in the frame image, the recognition module issues a photographing command, saves the photo, and starts the graphic analysis system.

If a plurality of colors are included in the scan area, as long as the number of consecutive pixel points of the plurality of colors in the frame image is equal to one-third or more of the number of images in one frame, then these colors need to be recognized. As the background color, the color that satisfies the background color requirement is set to the background color during scanning. During the scanning process, the photo command will not be issued unless the graphic recognition system recognizes the pixel with a large difference in background color.

The graphic analysis system further analyzes the color information and the pixel distribution of the identified pixel point set to determine whether the color area composed of the pixels is a stain or has been recorded by the system, if it is confirmed to be a stain and is not recorded by the system. Then record the center coordinate position of the stain here and save the current image. For the area scan of multiple background colors, the decision of the graphics analysis system for whether certain pixels are stains depends only on the background color of the area where the pixel is located, that is, if the color area is different from the background color of the area. When this is the case, the color area is judged as a stain.

For the case where some color regions appear between two or more background colors in the scan, the graphics analysis system compares the color of the pixel points of the region with all other background colors, as long as one of the background colors is If the difference is large, the color area is judged as a stain; if the color of the pixel of the area is the same as one or more of all other background colors, the color area is not judged as a stain.

During the actual scanning process, it may be found that there may be some patterns or textures in the scanning area itself. These patterns and textures are basically in the form of a line or a line, if the coordinates of the stain are found in A pattern similar to a line graph and a dot map can be formed on the image. For example, if the coordinates of the single pixel point of the obtained stain differ greatly from the center coordinate and are continuously distributed, the image is excluded.

The graphic analysis system finally saves the analyzed dirt photos, and collects all the saved dirt photos and coordinate information after the scanning is completed to obtain a scan area dirt map.

At the same time, the graphic analysis system outputs the pollution degree of the scanning area to the result output module according to the different ratio of the total area of each stain in the dirt map to the sum of the area of the photographs determined as the dirt photograph per frame.

Data storage module:

This module uses a separate memory or mobile phone memory, on the one hand to save all the photos taken by the scanning module during the scanning process, and on the other hand assists the graphic analysis system to read and write the dirt images of the saved graphics to the module. Given to save, after the module to be scanned finishes working, all the pictures and their coordinate information are sent to the number evaluation module.

The evaluation module uses the display of the mobile phone to output a preview of the entire scanning process. After the module to be scanned finishes working, the photo sent by the analysis module is displayed on the display screen, and the user can browse all the dirt photos. If it is found that it is not dirt, it can be manually deleted. After the user checks and confirms that the error is correct, the evaluation module can be based on the last The remaining photos determine the stain attachment of the medium to be tested.

The above is a preferred embodiment of the present invention, and it should be noted that those skilled in the art can also make several improvements and retouchings without departing from the principles of the present invention. It should be considered as the scope of protection of the present invention.

Industrial applicability

The embodiment of the invention detects and analyzes the dirt of the target object through graphic recognition, and is particularly suitable for the application scenario for determining the sanitary condition. In addition, the entire solution can be executed on terminals that are currently available on the market, so the cost is low and the penetration rate is high.

Claims

A method of detecting dirt, the method comprising:

Scanning step, scanning and imaging the medium to be tested according to the preset number of frames;

a step of identifying, according to a pixel distribution of each color of the imaging result of the current frame number, identifying a background color of the imaging result of the current frame number;

An analyzing step of determining, according to a pixel distribution of the non-background color of the imaging result of the current frame number, whether the imaging result of the current frame number has dirt;

In the photographing step, if it is determined that there is dirt in the imaging result of the current frame number, the imaging result of the current frame number is photographed, and the photographing result is obtained;

The evaluating step, after the end of the scanning step, evaluates the adhesion of the dirt of the medium to be tested based on all the obtained photographing results.
The method of claim 1 wherein said identifying step comprises:

Determining the pixel ratio of each color; wherein, the pixel ratio of a certain color = the number of consecutive pixels of the color / the number of all pixels in the imaging result of the current number of frames;

A color whose pixel ratio is greater than the first predetermined threshold is determined as the background color of the imaging result of the current number of frames.
The method of claim 2 wherein said analyzing step comprises:

Determining the sum of pixel proportions of non-background colors;

When the sum of the pixel ratios of the non-background colors is greater than the second predetermined threshold, it is determined that the imaging result of the current number of frames has a stain.
The method of claim 3 wherein said evaluating step comprises:

Determining the degree of soil adhesion of the medium to be tested according to the photographing result; the degree of stain adhesion=the number of pixels of the non-background color of all photographing results/the number of all the pixels of all photographing results;

The adhesion of the dirt of the medium to be tested is evaluated according to the degree of soil adhesion.
The method of claim 1 wherein said identifying step further comprises:

Configuring a coordinate system for all pixels of the imaging result of the current frame number;

Determining a graphic of each color according to coordinates corresponding to each color;

The color corresponding to the graphic that meets the preset rule is determined as the background color;

The preset rule includes: a circle and/or a line shape.
A device for detecting dirt includes: a scanning module, an identification module, an analysis module, a camera module, and an evaluation module; wherein

The scanning module is configured to scan and image the medium to be tested according to a preset number of frames;

The identification module is configured to identify a background color of the imaging result of the current frame number according to a pixel distribution of each color of the imaging result of the current frame number;

The analyzing module is configured to determine, according to a pixel distribution of the non-background color of the imaging result of the current frame number, whether the imaging result of the current frame number has dirt;

The photographing module is configured to: if it is determined that the imaging result of the current frame number is dirty, photographing the imaging result of the current frame number to obtain a photographing result;

The evaluation module is configured to evaluate the adhesion of the dirt of the medium to be tested according to all the obtained photographing results.
The apparatus of claim 6 wherein said identifying means comprises:

a first determining sub-module configured to determine a pixel ratio of each color; wherein, a pixel ratio of a certain color=the number of consecutive pixels of the color/the number of all pixels in the imaging result of the current frame number;

And a second determining submodule configured to determine a color whose pixel ratio is greater than the first preset threshold as a background color of the imaging result of the current frame number.
The apparatus of claim 7 wherein said analyzing means comprises:

a third determining submodule configured to determine a sum of pixel proportions of non-background colors;

And a fourth determining submodule configured to determine that the imaging result of the current frame number has a dirt when the sum of the pixel ratios of the non-background colors is greater than the second preset threshold.
The apparatus of claim 8 wherein said evaluating means comprises:

Processing the sub-module, determining, according to the photographing result, the dirt adhesion degree of the medium to be tested; the dirt adhesion degree=the number of pixels of the non-background color of all the photographing results/the number of all the pixels of all the photographing results;

The evaluation sub-module evaluates the adhesion of the dirt of the medium to be tested according to the degree of soil adhesion.
The apparatus of claim 6 wherein said identifying means further comprises:

a fifth determining submodule configured to configure a coordinate system for all pixels of the imaging result of the current frame number;

a sixth determining submodule configured to determine a graphic of each color according to coordinates corresponding to each color;

a seventh determining submodule configured to determine a color corresponding to the graphic that meets the preset rule as a background color;

The preset rule includes: a circle and/or a line shape.
A terminal comprising the apparatus for detecting dirt according to any one of claims 6-10.