CN101532829A - Method for detecting surface orange by analyzing intensity images of shadows formed on surface of object by light source - Google Patents
Method for detecting surface orange by analyzing intensity images of shadows formed on surface of object by light source Download PDFInfo
- Publication number
- CN101532829A CN101532829A CN200910061707A CN200910061707A CN101532829A CN 101532829 A CN101532829 A CN 101532829A CN 200910061707 A CN200910061707 A CN 200910061707A CN 200910061707 A CN200910061707 A CN 200910061707A CN 101532829 A CN101532829 A CN 101532829A
- Authority
- CN
- China
- Prior art keywords
- tangerine peel
- image
- light source
- dimension
- signal
- 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.)
- Granted
Links
Images
Landscapes
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention relates to a method for detecting surface orange by analyzing intensity images of shadows formed on the surface of an object by a light source, comprising the following steps: step 1: obtaining intensity images of shadows of the light source on the surface of the object detected; step 2: analyzing images by the detection algorithm to obtain the texture structure data of the surface orange; and step 3: converting the data obtained in the range of 0-99.9 according to the different detection objects and different calibration parameters to obtain the evaluation of the orange, the higher the value is, the less orange on the surface and better quality of the surface are. The invention has the beneficial effects of low requirements of hardware and detection environment, fast detection speed, and high-efficiency and mass detection of the texture structure of the orange on the surface of the object with high reflecting ability.
Description
Technical field
The invention belongs to field of coating and computer application field, especially high gloss material surface tangerine peel quality determining method is a kind of by analyzing light source detects surperficial tangerine peel at the luminance picture of image that body surface becomes method specifically.
Background technology
The coating surface outward appearance presents many semicircle shape projections, is called tangerine peel as the ripple of tangerine peel.
In early days, the detection mode of the tangerine peel quality of high gloss body surface is mainly the mode of human eye range estimation, but owing to there is not clear and definite evaluation criteria, and it is bigger influenced by observer's subjective feeling and personal experience, its result is objective inadequately.
People evaluated with roughmeter afterwards, utilized the varying degree physics tangerine peel as the visual assessment standard.But this mensuration can only be carried out in the laboratory, and the measured just mechanical outline on surface, non-optical profile.
At present, the detection of tangerine peel quality is mainly undertaken by laser tangerine peel scanner, this instrument adopts laser focusing on the testee surface, the optical profile structure on the light and shade mutation analysis surface by reflection ray.Exist equipment that hardware and testing environment are had relatively high expectations, and the shortcoming that involves great expense.
Orange peel on the body surface is comprehended the light and shade that causes surface reflection light and is changed.Human eyes can be seen the physical dimension between 0.1～10.0mm.When the rayed body surface, each position reflective light intensity difference of tangerine peel texture, wherein the local reflection ray at tangerine peel texture crest and trough is the strongest, and local reflection ray is the most weak on the slope of texture.The result of performance is light source and changes in the brightness of image that the surface becomes, and human eye has just formed the assessment of vision to tangerine peel by the contrast in high light district and non-light district.Therefore store light source in the monochrome information of the imaging of body surface and analyze, can read the tangerine peel data texturing of body surface.
Summary of the invention
The purpose of this invention is to provide a kind of with specific light source irradiating object surface and obtain its imaging on the surface, by the monochrome information of analyzing the light source image tangerine peel on surface is estimated, accurate, the easy detection of energy has the method for the tangerine peel structure of high reflectance body surface.
To achieve these goals, the method applied in the present invention is:
First step: obtain testee surface optics image: with light source irradiation high gloss body surface, utilize camera or sensor to obtain light source, image is stored in computing machine or the embedded system with the computer picture storage format at image that body surface becomes.
Second step: adopt detection algorithm that image is analyzed, draw the texture structure data of surperficial tangerine peel, its concrete grammar is:
The first step: image pre-service: extract light source image part in the image;
Second step: the light source image is partly carried out the one dimension luma samples;
The 3rd step: to the denoising of one dimension luminance signal: environmental noise and high frequency noise can be influential to evaluating result in testing process, therefore extract the luminance signal of energy accurate response body surface tangerine peel texture from two dimensional image, could guarantee the accuracy that tangerine peel detects.Adopt the EMD algorithm one dimension luminance signal to be resolved into the subsignal of different frequency in the method, wherein high frequency subsignal is the high frequency white noise, the low frequency subsignal is mainly system noise and comprises the influence to signal such as surround lighting and hardware device, can determine the subsignal of representative system noise according to physical device and situation, after removing white noise and system noise, remaining subsignal is synthetic, can obtain one dimension tangerine peel signal;
The 4th step: analyze one dimension tangerine peel signal: with the distance definition between two maximum points is a wavelength, at first find out all maximum points in the one dimension tangerine peel signal, calculate all wavelengths in the one dimension tangerine peel signal, its medium wavelength is long more, the tangerine peel structure that shows body surface is big more, otherwise it is then more little, wavelength is defined as long wave greater than 0.6mm, wavelength is defined as shortwave between 0.1mm-0.6mm, the quantity of statistics long wave and shortwave is calculated the long wave of this one dimension tangerine peel signal and the ratio of shortwave quantity;
The 5th step: the light source image is carried out repeatedly luma samples, go on foot the length Bob value that calculates in all one dimension tangerine peel signals according to the 3rd step-Di 4 in this step, and ask for their mean value Avg, to reduce the error of statistics, Avg promptly characterizes the data of surperficial tangerine peel;
Third step: the tangerine peel data are converted to the tangerine peel evaluation by parameter calibration.(a b), is transformed into mean value Avg in the corresponding scope, and this scope is consistent with the scope of the tangerine peel value of on-gauge plate to set one group of calibrating parameters at dissimilar detected objects.Calculate OP=Avg*a+b, OP is the evaluation to the measurand tangerine peel.
Definite method of calibrating parameters:
Because dissimilar object surfaces optical reflection performance differences, the characteristic element of their surface brightness images is also different to the reflection of the tangerine peel on surface.Therefore need set a calibrating parameters for dissimilar detected objects, make the tangerine peel on the correct evaluation object surface of detection algorithm energy, also their detected value be demarcated in the unified scope simultaneously.
The first step, at certain type detected object, selected n opens the on-gauge plate of the type.Known their tangerine peel value is respectively ST
1, ST
2... ST
n, the scope of its value is at 0-99.9.
Second goes on foot, this batch on-gauge plate is calculated according to the above-mentioned first step and second step their tangerine peel data Avg
1, Avg
2... Avg
n
The 3rd goes on foot, asks system of equations a[Avg
1, Avg
2... Avg
n]
T+ b=[ST
1, ST
2... ST
n] least square solution (a is b) as the calibrating parameters of such material surface.
The invention has the beneficial effects as follows: to hardware and testing environment require lowly, detection speed is fast, can be efficient, large batch of detection has the tangerine peel texture structure of high reflectance body surface.
Description of drawings
Fig. 1 is the original image of the present invention's body surface after the band light source irradiation.
Wherein dashed region is light source image locating area.
Fig. 2 is the pretreated image of the present invention.
Fig. 3 chooses the synoptic diagram of sweep trace at the light source image area for the present invention.
Wherein a black line in the bright band of top is a sweep trace of extraction.
Fig. 4 is the one dimension luminance signal figure of sweep trace of the present invention.
Fig. 5 carries out the synoptic diagram that wavelength-division is separated for the present invention adopts the EMD algorithm to sweep trace.
Wherein article one ripple signal from top to bottom is the high frequency white noise, and second to the 4th is the tangerine peel signal, and the 5th to the 7th is system noise.
The tangerine peel signal schematic representation that Fig. 6 extracts from the one dimension luminance signal for the present invention.
Embodiment
The present invention is described in further detail below in conjunction with drawings and Examples.
Concrete steps of the present invention are:
First step: Image Acquisition:
Light source and image acquisition element are fixed in composition one cover image capture device in the black box.There are two rectangular slot up and down on the black box surface, is an image acquisition hole between the slit, deploys image acquisition element in it, as CMOS/CDT camera lens or sensor.Built-in two light belts of being made up of some standard led light sources of black box are put scattering sheet between light belt and slit.When lighting the LED light belt, form two uniform rectangle light belts behind its light that sends process scattering sheet and the slit.This moment, rectangle light belt irradiating object surface utilized image capture device such as CMOS/CDT to gather light source at two-dimentional brightness image that body surface became.The distance of control collecting device and measurand makes the image of collection the most clear.
Second step: adopt detection algorithm that image is analyzed, draw surperficial tangerine peel texture structured value, its concrete grammar is:
1. image pre-service: comprise
A) positioned light source image:
With two-dimentional luminance picture with given threshold value 100 (can according to the actual conditions adjustment) binaryzation, obtain black white image, adopt Seed Filling algorithm to obtain all connected regions in black white image, wherein Zui Da two connected regions are the major part of light source image;
B) compute location zone:
The connected region proper transformation of being asked in a) is obtained locating area; Increase the height of connected region, make locating area can comprise complete image boundary information; Reduce the width of connected region, eliminate the image (Fig. 1) of end effect the result.
2. luma samples:
The light source image is carried out luma samples, the light source imagery zone laterally choose a sweep trace at a certain distance in two bright bands up and down, obtain m bar one dimension luminance signal B
j(j=1,2...m), the black line among Fig. 3 in the middle of the bright band is a sweep trace, and Fig. 4 is the one dimension luminance signal B of this sweep trace
j
3. extract the tangerine peel signal:
To the one dimension luminance signal B that chooses in 2
jDecompose, extract tangerine peel signal O after the denoising
j:
A) adopt EMD (empirical mode decomposition) algorithm that the one dimension luminance signal is decomposed into the subsignal (Fig. 5) of n bar different frequency, these subsignals once are numbered S from high to low by frequency
1... S
n
B) these subsignals are divided three classes:
S
1(article one subsignal) is the high frequency white noise;
S
2To S
i(
) be the tangerine peel signal;
S
iTo S
nBe system noise.
Extract tangerine peel signal S wherein
2To S
iSubsignal is also synthetic, promptly obtains one dimension tangerine peel signal O
j(Fig. 6):
4. add up:
A) statistics one dimension tangerine peel signal O
jIn the quantity of length ripple: with the ripple signal between adjacent two maximum points on the ripple signal or adjacent two minimum points as a complete ripple, definition distance therebetween is a wavelength, wavelength is reduced shortwave less than the ripple of 0.6mm, wavelength reduces long wave greater than 0.6mm less than the ripple of 10mm, the one dimension tangerine peel signal O after the statistics denoising
jIn the quantity LW of long wave
jWith shortwave quantity SW
j
B) calculate one dimension tangerine peel signal O
jIn LW
jAnd SW
jRatio R
j=LW
j/ SW
j
C) calculate length wave number amount ratio R in each tangerine peel signal
j(j=1, mean value Avg 2...m);
Third step: the tangerine peel data are converted to the tangerine peel evaluation by parameter calibration.
Mean value Avg is carried out linear transformation, and (a b) is converted to its value numerical value OP between 0 to 99.9, OP=a*Avg+b to the calibrating parameters by this detected object type.Numerical value OP after the conversion can characterize the tangerine peel texture structured value on testee surface, and mark is high more, and surperficial tangerine peel is more little, and quality is good more.
The content that is not described in detail in this instructions belongs to this area professional and technical personnel's known prior art.
Claims (8)
1, a kind ofly detect the method for surperficial tangerine peel by analyzing light source at the luminance picture of image that body surface becomes, the method that is adopted is:
First step: obtain testee surface optics image: with light source irradiation high gloss body surface, utilize camera or sensor to obtain light source, image is stored in computing machine or the embedded system with the computer picture storage format at image that body surface becomes;
Second step: adopt detection algorithm that image is analyzed, draw the texture structure data to surperficial tangerine peel, its concrete grammar is:
The first step: image pre-service: extract light source image part in the image;
Second step: the light source image is partly carried out the one dimension luma samples;
The 3rd step: to the denoising of one dimension luminance signal: adopt the EMD algorithm one dimension luminance signal to be resolved into the subsignal of different frequency, wherein high frequency subsignal is a high frequency white noise subsignal, the low frequency subsignal is the subsignal of system noise, after removing white noise and system noise, remaining subsignal is synthetic, obtain one dimension tangerine peel signal;
The 4th step: analyze one dimension tangerine peel signal: with the distance definition between two maximum points is a wavelength, at first find out all maximum points in the one dimension tangerine peel signal, calculate all wavelengths in the one dimension tangerine peel signal, its medium wavelength is long more, the tangerine peel structure that shows body surface is big more, otherwise then more little, the quantity of statistics long wave and shortwave is calculated the long wave of this one dimension tangerine peel signal and the ratio of shortwave quantity;
The 5th step: the light source image is carried out repeatedly luma samples, go on foot the length Bob value that calculates in all one dimension luminance signals according to the 3rd step-Di 4 in this step, and ask for their mean value Avg, as the data of reflection surface tangerine peel;
Third step: the tangerine peel data are converted to the tangerine peel evaluating data: set one group of calibrating parameters (a at dissimilar detected objects by parameter calibration, b), mean value Avg is transformed in the corresponding scope, by calculating the evaluating data OP=Avg*a+b of measurand tangerine peel, numerical value is high more, the surface tangerine peel is more little, and quality is good more.
2, as claimed in claim 1 by analyzing light source detects surperficial tangerine peel at the luminance picture of image that body surface becomes method, it is characterized in that: the concrete grammar that first step adopted is: light source and image acquisition element are fixed in composition one cover image capture device in the black box, there are two rectangular slot up and down on the black box surface, it between the slit image acquisition hole, deploy image acquisition element in it, built-in two light belts of forming by some standard led light sources of black box, between light belt and slit, put scattering sheet, when lighting the LED light belt, form two uniform rectangle light belts behind its light that sends process scattering sheet and the slit, this moment rectangle light belt irradiating object surface, utilize image acquisition element to gather light source at two-dimentional brightness image that body surface became, the distance of control collecting device and measurand makes the image of collection the most clear.
3, as claimed in claim 1ly detect the method for surperficial tangerine peel by analyzing light source at the luminance picture of image that body surface becomes, it is characterized in that: the concrete grammar of the first step comprises in second step:
A) positioned light source image:
Two-dimentional luminance picture with given threshold value 100 binaryzations, is obtained black white image, adopt the SeedFilling algorithm to obtain all connected regions in black white image, wherein Zui Da two connected regions are the major part of light source image;
B) compute location zone:
The connected region proper transformation of being asked in a) is obtained locating area; Increase the height of connected region, make locating area can comprise complete image boundary information; Reduce the width of connected region, eliminate the image of end effect the result.
4, as claimed in claim 1 by analyzing light source detects surperficial tangerine peel at the luminance picture of image that body surface becomes method, it is characterized in that: the concrete grammar in second step is in second step: the light source image is carried out luma samples, the light source imagery zone laterally choose a sweep trace at a certain distance in two bright bands up and down, obtain m bar one dimension luminance signal B
j(j=1,2...m).
5, as claimed in claim 1ly detect the method for surperficial tangerine peel by analyzing light source at the luminance picture of image that body surface becomes, it is characterized in that: the 3rd concrete grammar that goes on foot is in second step:
A) adopt the EMD algorithm with one dimension luminance signal B
jBe decomposed into the subsignal of n bar different frequency, these subsignals once are numbered S from high to low by frequency
1... S
n
B) these subsignals are divided three classes:
S
1Be the high frequency white noise;
S
2To S
iBe the tangerine peel signal, wherein
S
iTo S
nBe system noise;
Extract tangerine peel signal S wherein
2To S
iSubsignal is also synthetic, obtains one dimension tangerine peel signal O
j:
6, as claimed in claim 1ly detect the method for surperficial tangerine peel by analyzing light source at the luminance picture of image that body surface becomes, it is characterized in that: the 4th concrete grammar that goes on foot is in second step:
A) statistics one dimension tangerine peel signal O
jIn the quantity of length ripple: with the ripple signal between adjacent two maximum points on the ripple signal or adjacent two minimum points as a complete ripple, definition distance therebetween is a wavelength, wavelength is reduced shortwave less than the ripple of 0.6mm, wavelength reduces long wave greater than 0.6mm less than the ripple of 10mm, statistics one dimension tangerine peel signal O
jIn the quantity LW of long wave
jWith shortwave quantity SW
j
B) calculate one dimension tangerine peel signal O
jIn LW
jAnd SW
jRatio R
j=LW
j/ SW
j
7, detect the method for surperficial tangerine peel by analyzing light source at the luminance picture of image that body surface becomes as right 1 is described, it is characterized in that: the concrete grammar of the definite detected object calibrating parameters in third step is:
The first step, at the type of detected object, selected n opens the on-gauge plate of the type, known their tangerine peel value is respectively ST
1, ST
2... ST
n
Second goes on foot, this batch on-gauge plate is calculated according to the described first step and second step their tangerine peel data Avg
1, Avg
2... Avg
n
The 3rd goes on foot, asks system of equations a[Avg
1, Avg
2... Avg
n]
T+ b=[ST
1, ST
2... ST
n] least square solution (a is b) as the calibrating parameters of such material surface.
8, detect the method for surperficial tangerine peel by analyzing light source at the luminance picture of image that body surface becomes as right 1 is described, it is characterized in that: the concrete grammar that the mean value Avg of the length Bob value that is some sweep traces in the luminance picture to detected tangerine peel data in third step demarcates is:
(a b), is transformed into mean value Avg in the corresponding scope, and this scope is consistent with the tangerine peel value scope of on-gauge plate, calculates the evaluating data OP=Avg*a+b of measurand tangerine peel for the selected calibrating parameters of detected object.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100617072A CN101532829B (en) | 2009-04-24 | 2009-04-24 | Method for detecting surface orange by analyzing intensity images of shadows formed on surface of object by light source |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009100617072A CN101532829B (en) | 2009-04-24 | 2009-04-24 | Method for detecting surface orange by analyzing intensity images of shadows formed on surface of object by light source |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101532829A true CN101532829A (en) | 2009-09-16 |
| CN101532829B CN101532829B (en) | 2010-12-29 |
Family
ID=41103573
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009100617072A Expired - Fee Related CN101532829B (en) | 2009-04-24 | 2009-04-24 | Method for detecting surface orange by analyzing intensity images of shadows formed on surface of object by light source |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101532829B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103033162A (en) * | 2012-12-25 | 2013-04-10 | 浙江大学 | Detection method for silkworm cocoon corrugating |
| CN105158171A (en) * | 2015-06-08 | 2015-12-16 | 南京农业大学 | Spectral calibration method for crop nitrogen sensor |
| CN111156932A (en) * | 2020-03-10 | 2020-05-15 | 凌云光技术集团有限责任公司 | Mirror surface material roughness detection device |
| CN113031826A (en) * | 2021-04-15 | 2021-06-25 | 维沃移动通信有限公司 | Screen assembly and electronic equipment |
-
2009
- 2009-04-24 CN CN2009100617072A patent/CN101532829B/en not_active Expired - Fee Related
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103033162A (en) * | 2012-12-25 | 2013-04-10 | 浙江大学 | Detection method for silkworm cocoon corrugating |
| CN105158171A (en) * | 2015-06-08 | 2015-12-16 | 南京农业大学 | Spectral calibration method for crop nitrogen sensor |
| CN105158171B (en) * | 2015-06-08 | 2018-02-06 | 南京农业大学 | Crop nitrogen sensor spectrum calibration method |
| CN111156932A (en) * | 2020-03-10 | 2020-05-15 | 凌云光技术集团有限责任公司 | Mirror surface material roughness detection device |
| CN111156932B (en) * | 2020-03-10 | 2021-08-27 | 凌云光技术股份有限公司 | Mirror surface material roughness detection device |
| CN113031826A (en) * | 2021-04-15 | 2021-06-25 | 维沃移动通信有限公司 | Screen assembly and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101532829B (en) | 2010-12-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Goel et al. | HyperCam: hyperspectral imaging for ubiquitous computing applications | |
| CN109631787A (en) | The spot center detection method and deflection of bridge span image-type detection device of transmission-type target image | |
| CN101532829B (en) | Method for detecting surface orange by analyzing intensity images of shadows formed on surface of object by light source | |
| US20110063437A1 (en) | Distance estimating device, distance estimating method, program, integrated circuit, and camera | |
| US20200113441A1 (en) | Skin gloss measurement for quantitative estimation of skin gloss | |
| Xiong et al. | Color sensors and their applications based on real-time color image segmentation for cyber physical systems | |
| US11378521B2 (en) | Optical condition determination system and optical condition determination method | |
| US11159712B2 (en) | Range differentiators for auto-focusing in optical imaging systems | |
| RU2010152571A (en) | METHOD FOR AUTOMATIC EVALUATION OF SKIN AND / OR WRINKLE TEXTURE | |
| KR20170088259A (en) | Method and Apparatus FOR obtaining DEPTH IMAGE USING TOf(Time-of-flight) sensor | |
| Li et al. | Seasonal change of leaf and woody area profiles in a midlatitude deciduous forest canopy from classified dual-wavelength terrestrial lidar point clouds | |
| CN111415349A (en) | Method for detecting polyester filament yarn based on image processing technology | |
| KR20220147604A (en) | gesture recognition | |
| Zhang et al. | Quantitative evaluation of impact damage to apples using NIR hyperspectral imaging | |
| JP5590541B2 (en) | Freshness evaluation apparatus and freshness evaluation method | |
| CN106770287A (en) | A kind of one camera balanced type optical coherence tomography scanning means and method | |
| Chéné et al. | Multiscale analysis of depth images from natural scenes: Scaling in the depth of the woods | |
| WO2005087460A1 (en) | Apparatus, method and system for detecting the width and position of adhesives applied to a substrate | |
| Yang et al. | Optimized lighting method of applying shaped-function signal for increasing the dynamic range of LED-multispectral imaging system | |
| EP3561489A1 (en) | System and method for characterization of patterns marked on a fabric | |
| CN107730565B (en) | OCT image-based material intrinsic spectral feature extraction method | |
| Houser et al. | Shadow analysis via the C+ K Visioline: A technical note | |
| Kikuchi et al. | Age‐related changes in surface reflection, diffuse reflection, and subsurface scattering light of facial skin: luminance value measured by the system for the optical properties of facial skin | |
| Cruz-Guerrero et al. | Reflectance calibration with normalization correction in hyperspectral imaging | |
| JP4954038B2 (en) | Evaluation method of makeup scraps |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20101229 Termination date: 20130424 |