CN109242792A - A kind of white balance proofreading method based on white object - Google Patents
A kind of white balance proofreading method based on white object Download PDFInfo
- Publication number
- CN109242792A CN109242792A CN201810964247.3A CN201810964247A CN109242792A CN 109242792 A CN109242792 A CN 109242792A CN 201810964247 A CN201810964247 A CN 201810964247A CN 109242792 A CN109242792 A CN 109242792A
- Authority
- CN
- China
- Prior art keywords
- white
- image
- model
- white object
- white balance
- 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
- 238000000034 method Methods 0.000 title claims abstract description 37
- 230000001915 proofreading effect Effects 0.000 title claims abstract description 21
- 238000012549 training Methods 0.000 claims abstract description 31
- 238000005516 engineering process Methods 0.000 claims abstract description 10
- 239000000284 extract Substances 0.000 claims abstract description 9
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 7
- 238000012937 correction Methods 0.000 claims abstract description 4
- 238000012360 testing method Methods 0.000 claims description 13
- 239000000463 material Substances 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims description 2
- 230000009885 systemic effect Effects 0.000 claims description 2
- 238000013316 zoning Methods 0.000 claims description 2
- 238000004364 calculation method Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 15
- 238000003745 diagnosis Methods 0.000 description 13
- 239000003814 drug Substances 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 210000001835 viscera Anatomy 0.000 description 2
- 241001536352 Fraxinus americana Species 0.000 description 1
- 241000565357 Fraxinus nigra Species 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 240000005373 Panax quinquefolius Species 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 238000012850 discrimination method Methods 0.000 description 1
- 229940079593 drug Drugs 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000006641 stabilisation Effects 0.000 description 1
- 238000011105 stabilization Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/80—Geometric correction
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Radiology & Medical Imaging (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention discloses a kind of white balance proofreading method based on white object, comprising: using a large amount of white object samples obtained by constantly training study, establishes white object contrast model；It is captured on site using above-mentioned white object contrast model application, by identification image and extracts white object region in image, calculate its pixel mean value and colour gain values, the white balance check and correction of image is realized so as to adjust color；Realization principle of the present invention is simple, and resources costs are cheap, and operation time is quick, and white balance effect is ideal and accurate；The paper handkerchief identification technology that the present invention uses, based on convolutional neural networks algorithm, learn from big-sample data, gradually extracts the high-level characteristic of image, classify to feature, complete identification, the a degree of offset of paper handkerchief, dimensional variation and deformation can be coped with, guarantees the stronger separability of feature, there is ideal recognition effect to tagsort, identification is reduced to the dependence of external condition, while reducing the complexity of model.
Description
Technical field
The present invention relates to picture color process field more particularly to a kind of white balance proofreading methods based on white object.
Background technique
Traditional Chinese Medicine includes hoping, hearing, asking, cutting the four methods of diagnosis, and lingual diagnosis is the key content of observation.In tcm theory,
No matter human five internal organs' six internal organs how Fu Za pathological symptom, can intuitively, quickly by observe tongue picture judge disease property,
Shallow depth and the Sheng sorrow of qi and blood of patient's condition etc. are a kind of simple and effective medical aided diagnosis and discrimination method.Traditional observation, doctor
Life generally observes by the naked eye the features such as tongue form and the color of patient, is then diagnosed the illness according to career in medicine experience.Due to me
State is populous and Aging Problem is serious, and the patient populations to register the department of Chinese medicine daily are growing day by day, and supply falls short of demand by doctor；In addition, right
For patient, doctor is waited to see that examining is a cumbersome and time-consuming thing.
In order to solve above-mentioned conflict between doctors and patients, market has developed lingual diagnosis of a batch based on computer picture aid in treatment and produces
Product.The realization principle of the product is generally as follows: using patient's tongue as object, Image Acquisition is carried out, doctor is based on diagnostic imaging, this
Diagnosis efficiency is greatly improved.Since acquisition image is the capture apparatus by terminal, such as: mobile phone camera, the equipment do not have
There is the adaptability of human eye, is shot under different light, due to the disequilibrium of CCD output, the generation for causing colour reproduction to be distorted；
Therefore, often there is color difference in the acquisition image of above-mentioned lingual diagnosis product, and medical diagnosis is caused deviation also occur, can not be to patient
Disease, which is realized, targetedly treats and prevents.
Therefore, society need one kind is capable of the proofreading method of precise restoration acquisition image color.
Summary of the invention
The present invention aiming at the problems existing in the prior art, provides a kind of white balance proofreading method based on white object,
This method realization principle is simple, and resources costs are cheap, and operation time is quick, and white balance effect is ideal and accurate；This method uses
Paper handkerchief identification technology, be based on convolutional neural networks algorithm, learn from big-sample data, the high level for gradually extracting image is special
Sign, classifies to feature, completes identification, can cope with a degree of offset of paper handkerchief, dimensional variation and deformation, guarantee feature
Stronger separability has ideal recognition effect to tagsort, reduces identification to the dependence of external condition, reduces simultaneously
The complexity of model.
To achieve the above object, technical solution provided by the invention is as follows:
A kind of white balance proofreading method based on white object, comprising:
S1 is learnt by constantly training using a large amount of white object samples of acquisition, establishes white object contrast model；
S11, it is a large amount of to collect white object image, establish object sample database；
S111, a large amount of still images for collecting white object；
The still image of above-mentioned collection is carried out gray proces, calculates the gray scale of each pixel of each still image by S112
Value, makes above-mentioned image that black-white-gray state all be presented；
Still image sample after gray proces is divided into two class of test image and training image, stored by S113
On the server, white object sample database is established.
S12 establishes model, carries out model training using above-mentioned sample, realizes white object automatic identification.
S121, system establish model, the training image of above-mentioned white object sample are transferred to model repetition training, automatically
Identify the white object of image；
Whether S122, judgment models frequency of training reach the threshold value of systemic presupposition, if it is not, then S121 is gone to, if so,
Then go to S123；
S123, model training stop, and by the calculating of loss function, obtain model recognition accuracy；
S124 judges whether reach a certain threshold value by the above-mentioned accuracy rate obtained, if it is not, then readjusting the class of sample
Other information, if it is, going to S125；
The test image of model and white object sample database is carried out trial operation test by S125, and test accuracy rate reaches
Model application is captured on site after to preset threshold value.
S2 is captured on site using above-mentioned white object contrast model application, by identification image and is extracted in image
White object region calculates its pixel mean value and colour gain values, and the white balance check and correction of image is realized so as to adjust color.
S21, model application are captured on site, save the group photo image of user's tongue and white object；
S22 identifies the white object of user images, by above-mentioned white object contrast model, using identification technology, base
Learn from sample data in algorithm, gradually extract the high-level characteristic of image, classify to feature, completes identification；
S23 extracts the white object region of user images and cuts to the zone boundary, retains region middle part
Point；
S24, the mean value of each pixel in zoning middle section, the color value as white object；
S25 compares object color value and reference white color value, obtains colour gain values between the two；
S26 is based on above-mentioned yield value, carries out whole color to user images and adjusts；
S27, output have carried out the user images of color adjustment.
Further, calculated in the step S112 gray value of each pixel with formula add for gray proces
Weight average method formula:
F (i, j)=0.30R (i, j)+0.59G (i, j)+0.11B (i, j)
Wherein, i, j represent a pixel in the position of two-dimensional space vector, it may be assumed that the i-th row, jth column.
Further, the sample class information in the step S124 includes the material and size of object.
Further, the algorithm that identification technology is based in the step S22 is convolutional neural networks algorithm.
Further, the threshold value of model training number is 200,000 times in the step S122, model in the step S124
Training accuracy rate threshold value is 80%, and model measurement accuracy rate threshold value is 80% in the step S125.
Further, the white object is white paper towels.
Compared with prior art, the method for the present invention realization principle is simple, and resources costs are cheap, and operation time is quick, Bai Ping
The effect that weighs is ideal and accurate；The paper handkerchief identification technology that this method uses is based on convolutional neural networks algorithm, from big-sample data
Study, gradually extract image high-level characteristic, classify to feature, complete identification, can cope with paper handkerchief it is a degree of offset,
Dimensional variation and deformation guarantee the stronger separability of feature, have ideal recognition effect to tagsort, reduce identification pair
The dependence of external condition, while reducing the complexity of model.
Detailed description of the invention
Fig. 1: for the specific flow chart of white balance proofreading method step S1 of the present invention；
Fig. 2: for the specific flow chart of white balance proofreading method step S2 of the present invention.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiments of the present invention will be described in further detail.Implement below
Example is not intended to limit the scope of the invention for illustrating the present invention.
Image color is acquired in order to precise restoration, the white balance check and correction based on white object that the present invention provides a kind of
Method.White paper towels sample as a comparison is used in embodiment.
Referring to Fig.1, system largely collects white paper towels image, establishes paper handkerchief sample database.Model is established, based on above-mentioned
White paper towels automatic identification is realized in sample repetition training；By the calculating of loss function, if model recognition accuracy reaches a certain
Threshold values, such as: 80%, then model application is captured on site.
Referring to Fig. 2, in the field, user saves the group photo of my tongue and white paper towels by lingual diagnosis terminal.Based on upper
Model is stated, system carries out paper handkerchief identification to user images.Based on recognition result, system extracts the paper handkerchief region area Bing Yigai of image
The 10% of domain sizes respectively cuts boundary, retains the middle section in region.Based on above-mentioned middle section, system-computed is each
The average value of pixel, the color value as paper handkerchief region.Based on above-mentioned color value, system is the color value and reference white face
Color value compares, and obtains colour gain values between the two.Based on above-mentioned yield value, system recalculates the face of user images
Color value realizes user images color white balance effect.
Color characteristic of the present embodiment based on white paper towels provides white object of reference for lingual diagnosis image, and combined standard is white
The comparison of color obtains colour gain values between the two；Based on yield value, lingual diagnosis image color white balance effect is realized.This hair
Bright realization principle is simple, and resources costs are cheap, and operation time is quick, and white balance effect is ideal and accurate.It is based on white balance in the past
The proofreading method of technology: image is carried out piecemeal first；Then the average value based on piecemeal and accumulated value detection image is white
Color dot；Then final white reference point is determined；Be then based in the average value and reference point of above-mentioned reference point brightness maxima into
Row comparison, obtains yield value；It is finally based on yield value, color adjustment is carried out to image.Compared with the present invention, previous methods exist
Following deficiency: realization principle is complicated, and resources costs are high, computationally intensive, and working efficiency is low；And image has color difference originally, than
Pair brightness maxima reference point be also not necessarily reference white, also mean that the adjustment of next color of image still remains
The possibility of color difference, white balance effect inaccuracy.The paper handkerchief identification technology that the present invention uses is based on convolutional neural networks algorithm, from
Learn in big-sample data, gradually extract the high-level characteristic of image, classify to feature, completes identification；Paper handkerchief one can be coped with
Determine offset, dimensional variation and the deformation of degree, guarantee the stronger separability of feature, there is ideal identification effect to tagsort
Fruit reduces identification to the dependence of external condition, while reducing the complexity of model.
In embodiment, specific implementation method is as follows:
S1: it is a large amount of to collect white paper towels image, establish paper handkerchief sample database.
S1.1: a large amount of to collect white paper towels image.
By modes such as webpage captures, system largely collects the still image of white paper towels.
S1.2: image is carried out gray proces.
Since color image is made of multiple pixels, and each pixel is indicated by tri- values of RGB；To image into
Row gray proces can't influence the texture feature information of image, and each pixel only needs a gray value that can indicate,
Substantially increase image processing efficiency.It is gray proces weighted mean method formula below:
F (i, j)=0.30R (i, j)+0.59G (i, j)+0.11B (i, j)
Wherein, i, j represent a pixel in the position of two-dimensional space vector, it may be assumed that the i-th row, jth column.According to above-mentioned public affairs
Formula calculates the gray value of each pixel of each still image, and value range is 0-255, makes above-mentioned image that black and white ash shape all be presented
State.
S1.3: it on the server image storage, completes paper handkerchief sample database and establishes.
Still image passes through gray proces, becomes grayscale image.Based on total number of images amount, system divides the image into two classes, and one
Class is training image, and one kind is test image.Wherein, training image occupies the 90% of total quantity, when being mainly used for model training
It uses；Test image occupies the 10% of total quantity, is mainly used for model training and finishes, and when trial operation uses.Above-mentioned image is deposited
On local server, the foundation of paper handkerchief sample database is finished for storage.
S2: establishing model, realizes white paper towels automatic identification.
System establishes model, gives model repetition training above-mentioned paper handkerchief sample delivery.And model training mainly uses convolution
Neural network method learns from big-sample data, and the advanced features exported using different convolutional layers classify to feature,
To complete white paper towels automatic identification.When model and the number of paper handkerchief sample database training image repetition training reach system
Threshold values is set, such as: 200,000 times, model training stops.According to the calculating of loss function, model recognition accuracy is obtained.If accurate
Rate reaches a certain threshold values, such as: 80%, model is considered as ideal, can carry out trial operation test with paper handkerchief sample database test image；
Otherwise, model is considered as undesirable, readjusts the classification information of paper handkerchief, such as: material, size continue and paper handkerchief sample data
Library training image repetition training.
S3: model application is captured on site, carries out paper handkerchief identification to user images.
It trains and finishes when model and paper handkerchief sample database test image, and accuracy rate reaches a certain threshold values, such as: 80%,
It can be applied to live candid photograph.In the field, user saves the group photo of my tongue and white paper towels by lingual diagnosis terminal.Based on upper
Model is stated, system carries out paper handkerchief identification to user images.
S4: it is based on model recognition result, the paper handkerchief region of user images is extracted and calculates its color.
Based on above-mentioned model, system has realized that white paper towels identify to user images.According to recognition result, system, which is extracted, to be used
The paper handkerchief region of family image.It is white due to extracting field color, and there are also other colors for area peripheral edge, lead to zone boundary position
The white set suffers from the transmission or reflections affect of periphery color；Such as: area peripheral edge color is red, zone boundary position
White can then mix part red.In order to ensure the stabilization of said extracted region white, system is with the 10% of area size
Benchmark respectively cuts zone boundary, retains the middle section in region.Based on above-mentioned middle section, each picture of system-computed
The average value of vegetarian refreshments color, the color value as region, it may be assumed that the color value of white paper towels in user images.
S5: it is based on paper handkerchief color, it is compared with reference white, obtains color gain between the two.Based on being
The calculating united to each pixel color mean value in region middle section is extracted, has obtained the specific color value of paper handkerchief.Due to current paper
The color value of towel has color difference, and in order to ensure the white balance effect of user images, system is current paper handkerchief color value and standard white
Color color value is compared, and obtains colour gain values between the two.And under RGB color mode, reference white is R=G=B
=255.It is the formula for calculating color gain below:
Rgain=Ymax/Ravew
Ggain=Ymax/Gavew
Bgain=Ymax/Bavew
By above-mentioned formula, the yield value of user images paper handkerchief and reference white each Color Channel between the two is obtained.Its
In, Rgain, Ggain, Bgain represent the yield value of above-mentioned Color Channel each between the two；Ymax represents each color of reference white
The color value in channel；Ravew, Gavew, Bavew represent the color value of each Color Channel of user images paper handkerchief.
S6: being based on color gain, carries out integral color adjustment to user images.
Comparison based on user images paper handkerchief color value Yu reference white color value, has obtained color gain between the two
Value.Based on above-mentioned yield value, system recalculates the color value of each Color Channel of user images, and whole color is carried out to image
Image color white balance effect is realized in adjustment.It is the formula for calculating each Color Channel color value of image below:
R '=R × Rgain
G '=G × Ggain
B '=B × Bgain
By above-mentioned formula, the color value of each Color Channel of user images is obtained.Wherein, R', G', B' representative image color
The color value of each Color Channel after adjustment；R, the color value of each Color Channel before G, B representative image color adjust.
S7: the user images of output color adjustment.
Colour gain values based on image paper handkerchief and reference white, system carry out whole color tune to user images
It is whole, and exported by lingual diagnosis terminal, for diagnosis.
The present invention provides a kind of white balance proofreading method based on white object, has the advantage that the method for the present invention reality
Existing principle is simple, and resources costs are cheap, and operation time is quick, and white balance effect is ideal and accurate；The paper handkerchief that this method uses is known
Other technology is based on convolutional neural networks algorithm, learns from big-sample data, the high-level characteristic of image is gradually extracted, to feature
Classify, complete identification, a degree of offset of paper handkerchief, dimensional variation and deformation can be coped with, guaranteeing that feature is stronger can
Divide property, there is ideal recognition effect to tagsort, reduces identification to the dependence of external condition, while reducing answering for model
Miscellaneous degree.
The foregoing is merely presently preferred embodiments of the present invention, is not intended to restrict the invention, it is all in spirit of the invention and
In principle, any modification, equivalent replacement, improvement and so on be should all be included in the protection scope of the present invention.
Claims (10)
1. a kind of white balance proofreading method based on white object characterized by comprising
S1 is learnt by constantly training using a large amount of white object samples of acquisition, establishes white object contrast model；
S2 is captured on site using above-mentioned white object contrast model application, is passed through identification image and is extracted white in image
Object area calculates its pixel mean value and colour gain values, and the white balance check and correction of image is realized so as to adjust color.
2. white balance proofreading method according to claim 1, which is characterized in that the step S1 is specifically included:
S11, it is a large amount of to collect white object image, establish object sample database；
S12 establishes model, carries out model training using above-mentioned sample, realizes white object automatic identification.
3. white balance proofreading method according to claim 2, which is characterized in that the step S11 is specifically included:
S111, a large amount of still images for collecting white object；
The still image of above-mentioned collection is carried out gray proces, calculates the gray value of each pixel of each still image, make by S112
Black-white-gray state is all presented in above-mentioned image；
Still image sample after gray proces is divided into two class of test image and training image, is stored in clothes by S113
It is engaged on device, establishes white object sample database.
4. white balance proofreading method according to claim 3, which is characterized in that calculate each pixel in the step S112
The formula that the gray value of point is used is gray proces weighted mean method formula:
F (i, j)=0.30R (i, j)+0.59G (i, j)+0.11B (i, j)
Wherein, i, j represent a pixel in the position of two-dimensional space vector, it may be assumed that the i-th row, jth column.
5. white balance proofreading method according to claim 2, which is characterized in that the step S12 is specifically included:
S121, system establish model, and the training image of above-mentioned white object sample is transferred to model repetition training, automatic identification
The white object of image；
Whether S122, judgment models frequency of training reach the threshold value of systemic presupposition, if it is not, then S121 is gone to, if it is, turning
To S123；
S123, model training stop, and by the calculating of loss function, obtain model recognition accuracy；
S124 judges whether reach a certain threshold value by the above-mentioned accuracy rate obtained, if it is not, then readjusting the classification letter of sample
Breath, if it is, going to S125；
The test image of model and white object sample database is carried out trial operation test by S125, and test accuracy rate reaches pre-
If threshold value after model application is on site captured.
6. white balance proofreading method according to claim 5, which is characterized in that the sample class letter in the step S124
Breath includes the material and size of object.
7. white balance proofreading method according to claim 1, which is characterized in that the step S2 is specifically included:
S21, model application are captured on site, save the group photo image of user's tongue and white object；
S22 identifies the white object of user images, by above-mentioned white object contrast model, using identification technology, based on calculation
Method learns from sample data, gradually extracts the high-level characteristic of image, classifies to feature, completes identification；
S23 extracts the white object region of user images and cuts to the zone boundary, retains region middle section；
S24, the mean value of each pixel in zoning middle section, the color value as white object；
S25 compares object color value and reference white color value, obtains colour gain values between the two；
S26 is based on above-mentioned yield value, carries out whole color to user images and adjusts；
S27, output have carried out the user images of color adjustment.
8. white balance proofreading method according to claim 7, which is characterized in that identification technology institute base in the step S22
In algorithm be convolutional neural networks algorithm.
9. white balance proofreading method according to claim 5, which is characterized in that model training number in the step S122
Threshold value be 200,000 times, model training accuracy rate threshold value is 80% in the step S124, model measurement in the step S125
Accuracy rate threshold value is 80%.
10. claim 1 to claim 9 it is any as described in white balance proofreading method, which is characterized in that the white object
For white paper towels.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810964247.3A CN109242792B (en) | 2018-08-23 | 2018-08-23 | White balance correction method based on white object |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810964247.3A CN109242792B (en) | 2018-08-23 | 2018-08-23 | White balance correction method based on white object |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109242792A true CN109242792A (en) | 2019-01-18 |
| CN109242792B CN109242792B (en) | 2020-11-17 |
Family
ID=65068283
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810964247.3A Active CN109242792B (en) | 2018-08-23 | 2018-08-23 | White balance correction method based on white object |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109242792B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110740306A (en) * | 2019-10-24 | 2020-01-31 | 深圳市视特易智能科技有限公司 | White balance statistical correction template and method for color camera |
| CN112333437A (en) * | 2020-09-21 | 2021-02-05 | 宁波萨瑞通讯有限公司 | AI camera debugging parameter generator |
| CN112532960A (en) * | 2020-12-18 | 2021-03-19 | Oppo（重庆）智能科技有限公司 | White balance synchronization method and device, electronic equipment and storage medium |
| CN114071106A (en) * | 2020-08-10 | 2022-02-18 | 合肥君正科技有限公司 | Cold-start rapid white balance method for low-power-consumption equipment |
Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1079255A2 (en) * | 1999-08-23 | 2001-02-28 | Olympus Optical Co., Ltd. | Light source device for endoscope using dmd |
| CN1678081A (en) * | 2004-03-30 | 2005-10-05 | 豪威科技有限公司 | Method and apparatus for automatic white balance |
| WO2013118065A1 (en) * | 2012-02-06 | 2013-08-15 | Biooptico Ab | Camera arrangement and image processing method for quantifying tissue structure and degeneration |
| CN103957396A (en) * | 2014-05-14 | 2014-07-30 | 姚杰 | Image processing method and device used when tongue diagnosis is conducted with intelligent device and equipment |
| CN104658003A (en) * | 2015-03-16 | 2015-05-27 | 北京理工大学 | Tongue image segmentation method and device |
| CN104856680A (en) * | 2015-05-11 | 2015-08-26 | 深圳贝申医疗技术有限公司 | Automatic detection method and system for neonatal jaundice |
| CN105738364A (en) * | 2015-12-28 | 2016-07-06 | 清华大学深圳研究生院 | Measuring method and measuring apparatus of growth degree of algae on silicone rubber surface on the basis of image processing |
| CN106295139A (en) * | 2016-07-29 | 2017-01-04 | 姹ゅ钩 | A kind of tongue body autodiagnosis health cloud service system based on degree of depth convolutional neural networks |
| CN106339719A (en) * | 2016-08-22 | 2017-01-18 | 微梦创科网络科技（中国）有限公司 | Image identification method and image identification device |
| CN106412547A (en) * | 2016-08-29 | 2017-02-15 | 厦门美图之家科技有限公司 | Image white balance method and device based on convolutional neural network, and computing device |
| CN107507250A (en) * | 2017-06-02 | 2017-12-22 | 北京工业大学 | A kind of complexion tongue color image color correction method based on convolutional neural networks |
| CN107578390A (en) * | 2017-09-14 | 2018-01-12 | 长沙全度影像科技有限公司 | A kind of method and device that image white balance correction is carried out using neutral net |
| CN108205671A (en) * | 2016-12-16 | 2018-06-26 | 浙江宇视科技有限公司 | Image processing method and device |
| CN109273071A (en) * | 2018-08-23 | 2019-01-25 | 广东数相智能科技有限公司 | A method of establishing white object contrast model |
-
2018
- 2018-08-23 CN CN201810964247.3A patent/CN109242792B/en active Active
Patent Citations (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1079255A2 (en) * | 1999-08-23 | 2001-02-28 | Olympus Optical Co., Ltd. | Light source device for endoscope using dmd |
| CN1678081A (en) * | 2004-03-30 | 2005-10-05 | 豪威科技有限公司 | Method and apparatus for automatic white balance |
| WO2013118065A1 (en) * | 2012-02-06 | 2013-08-15 | Biooptico Ab | Camera arrangement and image processing method for quantifying tissue structure and degeneration |
| CN103957396A (en) * | 2014-05-14 | 2014-07-30 | 姚杰 | Image processing method and device used when tongue diagnosis is conducted with intelligent device and equipment |
| CN104658003A (en) * | 2015-03-16 | 2015-05-27 | 北京理工大学 | Tongue image segmentation method and device |
| CN104856680A (en) * | 2015-05-11 | 2015-08-26 | 深圳贝申医疗技术有限公司 | Automatic detection method and system for neonatal jaundice |
| CN105738364A (en) * | 2015-12-28 | 2016-07-06 | 清华大学深圳研究生院 | Measuring method and measuring apparatus of growth degree of algae on silicone rubber surface on the basis of image processing |
| CN106295139A (en) * | 2016-07-29 | 2017-01-04 | 姹ゅ钩 | A kind of tongue body autodiagnosis health cloud service system based on degree of depth convolutional neural networks |
| CN106339719A (en) * | 2016-08-22 | 2017-01-18 | 微梦创科网络科技（中国）有限公司 | Image identification method and image identification device |
| CN106412547A (en) * | 2016-08-29 | 2017-02-15 | 厦门美图之家科技有限公司 | Image white balance method and device based on convolutional neural network, and computing device |
| CN108205671A (en) * | 2016-12-16 | 2018-06-26 | 浙江宇视科技有限公司 | Image processing method and device |
| CN107507250A (en) * | 2017-06-02 | 2017-12-22 | 北京工业大学 | A kind of complexion tongue color image color correction method based on convolutional neural networks |
| CN107578390A (en) * | 2017-09-14 | 2018-01-12 | 长沙全度影像科技有限公司 | A kind of method and device that image white balance correction is carried out using neutral net |
| CN109273071A (en) * | 2018-08-23 | 2019-01-25 | 广东数相智能科技有限公司 | A method of establishing white object contrast model |
Non-Patent Citations (1)
| Title |
|---|
| 宋坤: "基于舌像的移动医疗诊断平台的设计与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110740306A (en) * | 2019-10-24 | 2020-01-31 | 深圳市视特易智能科技有限公司 | White balance statistical correction template and method for color camera |
| CN110740306B (en) * | 2019-10-24 | 2021-05-11 | 深圳市视特易智能科技有限公司 | Color white balance statistical correction method |
| CN114071106A (en) * | 2020-08-10 | 2022-02-18 | 合肥君正科技有限公司 | Cold-start rapid white balance method for low-power-consumption equipment |
| CN114071106B (en) * | 2020-08-10 | 2023-07-04 | 合肥君正科技有限公司 | Cold start fast white balance method for low-power-consumption equipment |
| CN112333437A (en) * | 2020-09-21 | 2021-02-05 | 宁波萨瑞通讯有限公司 | AI camera debugging parameter generator |
| CN112333437B (en) * | 2020-09-21 | 2022-05-31 | 宁波萨瑞通讯有限公司 | AI camera debugging parameter generator |
| CN112532960A (en) * | 2020-12-18 | 2021-03-19 | Oppo（重庆）智能科技有限公司 | White balance synchronization method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109242792B (en) | 2020-11-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109242792A (en) | A kind of white balance proofreading method based on white object | |
| US10115191B2 (en) | Information processing apparatus, information processing system, information processing method, program, and recording medium | |
| EP1450287B1 (en) | Method of extracting region of interest from tongue image and health monitoring method and apparatus using the tongue image | |
| CN104680524B (en) | A kind of leafy vegetable disease screening method | |
| CN108389207A (en) | A kind of the tooth disease diagnosing method, diagnostic device and intelligent image harvester | |
| CN106023151B (en) | Tongue object detection method under a kind of open environment | |
| WO2016179981A1 (en) | Automatic detection method and system for neonatal jaundice | |
| CN109635871A (en) | A kind of capsule endoscope image classification method based on multi-feature fusion | |
| CN109785943A (en) | A kind of monitoring of pathology and diagnostic message processing system and method | |
| CN110738655A (en) | Image report generation method, device, terminal and storage medium | |
| Engelke et al. | Visual attention modelling for subjective image quality databases | |
| CN109241963B (en) | Adaboost machine learning-based intelligent identification method for bleeding point in capsule gastroscope image | |
| CN107066779A (en) | Self-service body parameter inspection method and system based on biological specimen color | |
| CN115965607A (en) | Intelligent traditional Chinese medicine tongue diagnosis auxiliary analysis system | |
| JP2022137109A (en) | Image display system and image display method | |
| CN113139937A (en) | Digestive tract endoscope video image identification method based on deep learning | |
| CN116884073A (en) | Image recognition-based drinking detection method, device, medium and equipment | |
| CN109273071A (en) | A method of establishing white object contrast model | |
| CN115844436A (en) | CT scanning scheme self-adaptive formulation method based on computer vision | |
| CN113052813A (en) | Dyeing method based on StrainNet | |
| KR101566618B1 (en) | Method and apparatus for estimating hyperemia grade of eye | |
| Zhou et al. | Wireless capsule endoscopy video automatic segmentation | |
| Pan et al. | Bleeding detection from wireless capsule endoscopy images using improved euler distance in CIELab | |
| CN107689031B (en) | Color restoration method based on illumination compensation in tongue picture analysis | |
| Manchalwar et al. | Histogram of oriented gradient based automatic detection of eye diseases |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |