CN105096265A - Colour offset classification method based on histogram features - Google Patents
Colour offset classification method based on histogram features Download PDFInfo
- Publication number
- CN105096265A CN105096265A CN201510305816.XA CN201510305816A CN105096265A CN 105096265 A CN105096265 A CN 105096265A CN 201510305816 A CN201510305816 A CN 201510305816A CN 105096265 A CN105096265 A CN 105096265A
- Authority
- CN
- China
- Prior art keywords
- histogram
- type
- damaged
- channel
- colour cast
- 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
Abstract
The invention provides a colour offset classification method based on histogram features. An intelligent colour offset classification and correction method is provided; colour offsets are classified based on the histogram features at first; the colour offsets are corrected by using different methods according to a classification result; the colour offset classification method has complete self-adaptability no matter classification or correction; and manual intervention is unnecessary.
Description
Technical field
The present invention relates to Computer Image Processing field, particularly relate to a kind of colour cast classification processing method based on histogram feature for the enhancing of Computer Image Processing and image or recovery.
Background technology
Bearing calibration for colour cast is a lot, as the tone saturation degree based on rgb space histogrammic tone homogenizing, HSV space adjust, the pointwise correction supposed based on overall grey balance, darkChannle etc. based on RETINEX.But image is ever-changing, no matter which kind of method, all can not ensure that it is all optimum for adapting to all colour cast image conditions or under any circumstance correcting result, has his own strong points often.If Fig. 1 (a), Fig. 1 (b), Fig. 1 (c), Fig. 2 (a), Fig. 2 (b), Fig. 2 (c) are for subchannel matlab method and grey balance pointwise correction method, contrast test is carried out: can see from group contrast of two Fig. 1 (a), Fig. 1 (b), Fig. 1 (c) with different colour cast image, the correcting image utilizing grey balance pointwise correction method to obtain in first group is partially red, and the image color more true nature after the correction of tone homogenizing method; Just in time contrary in second picture group 2 (a), Fig. 2 (b), Fig. 2 (c), the effect of tone homogenizing method is far inferior to the latter, adopt the figure after the process of tone homogenizing, the inclined purple of its color, and the figure color more true nature after the process of grey balance pointwise correction.Example can prove that disposal route is different intuitively, and the effect image otherness of its process, how selecting suitable method to process image, is the problem that will solve required for the present invention.
Summary of the invention
For above-mentioned technological deficiency, the present invention proposes a kind of colour cast classification processing method based on histogram feature.
In order to solve the problems of the technologies described above, technical scheme of the present invention is as follows:
Based on a colour cast classification processing method for histogram feature, comprise the steps:
1.1) channel information amount is calculated:
1.2) according to channel information amount, image is divided into single channel type of damaged, binary channels type of damaged, channel offset type three class;
1.3) corresponding pattern colour bias correcting method is selected respectively according to single channel type of damaged, binary channels type of damaged, channel offset type three class.
Further, calculate channel information amount and calculate road histogram width for amounting to, the method that histogram width calculates refers to statistical pixel number from histogram both sides respectively, until reach total pixel 5%, then this place is distributing edge, and border, both sides difference is histogram distribution width.
Further, described single channel type of damaged is under rgb space, a certain channel histogram width fewer than other two passage 100 time, be considered as single channel impaired.
Further, described binary channels is impaired is under rgb space, a certain channel histogram width more than other two passage 100 time, be considered as binary channels impaired.
Further, described channel offset type is under rgb space, and histogram width difference is less than 100.
Further, single channel type of damaged, binary channels type of damaged all adopt grey balance color cast correction method, adopt the generating mode of different luminance graphs according to single channel type of damaged, binary channels type of damaged.
Further, when carrying out color cast correction for channel offset type employing subchannel matlab method, RETINEX bearing calibration.
Beneficial effect of the present invention is: the colour cast classification and the correction processing method that provide a kind of intelligence, first based on histogram feature, colour cast is classified, make according to classification results differently to correct it, no matter classify or correct, all there is adaptivity completely, without the need to manual intervention.
Accompanying drawing explanation
Fig. 1 (a) is original graph;
Fig. 1 (b) be by the process of Fig. 1 (a) tone homogenizing after figure;
Fig. 1 (c) be by the process of Fig. 1 (a) grey balance pointwise correction after figure;
Fig. 2 (a) is original graph;
Fig. 2 (b) be by the process of Fig. 2 (a) tone homogenizing after figure;
Fig. 2 (c) be by the process of Fig. 2 (a) grey balance pointwise correction after figure;
Fig. 3 is process flow diagram of the present invention;
Fig. 4 is the instance graph of channel offset type colour cast;
Fig. 4 (a) be by process of the present invention after figure;
Fig. 4 (b) be by the process of conventional process mode after figure;
Fig. 5 is the instance graph of single channel type of damaged colour cast;
Fig. 5 (a) be by process of the present invention after figure;
Fig. 5 (b) be by the process of conventional process mode after figure;
Fig. 6 is the instance graph of binary channels type of damaged colour cast;
Fig. 6 (a) be by process of the present invention after figure;
Fig. 6 (b) be by the process of conventional process mode after figure.
Embodiment
Below in conjunction with the drawings and specific embodiments, the present invention is described further.
As shown in Figure 3, based on histogram feature, colour cast is classified, making according to classification results differently to correct it, no matter classifying or correcting, all there is adaptivity completely, without the need to manual intervention.
Channel information amount: weighing main dependence standard is channel histogram width, the distribution range that namely pixel is main in histogram.Distribution range is larger, and to represent the contrast larger layers of this passage pixel time clearly demarcated, then quantity of information is larger; On the contrary, then contrast as little in distribution range little image shortage hierarchical information amount is little.The method of channel histogram width calculation refers to statistical pixel number from image histogram both sides respectively, until reach total pixel 5%, then this place is distributing edge, and border, both sides difference is histogram distribution width.
Colour cast is classified: the main subchannel quantity of information that relies on is classified, and image can be divided into single channel type of damaged, binary channels type of damaged, channel offset type three class.
Single channel is impaired: under referring to rgb space, only has an independent channel information to be subject to major injury, causes this channel information amount seriously lower than the situation of other two passage.Mainly interchannel quantity of information is poor for the criterion of such situation, and nisi information magnitude.A certain channel histogram width fewer than other two passage 100 time, be considered as single channel impaired.
Binary channels is impaired: under referring to rgb space, has the information of two passages to be subject to major injury, causes the situation that the quantity of information of they and the 3rd passage has a long way to go.A certain channel histogram width more than other two passage 100 time, be considered as binary channels impaired.
Channel offset type: under referring to rgb space, three-channel quantity of information is substantially suitable, and histogram width difference is less than 100, but their distributing positions on the histogram differ greatly, and can be considered channel offset.
The disposal route corresponding according to colour cast categorizing selection: channel offset type colour cast is more common colour cast type, the type each channel information amount is suitable, do not need the quantity of information across passage to mend, be therefore suitable for common color misregistration correction method, as subchannel matlab, RETINEX correct.Passage type of damaged colour cast, need interchannel information to supplement, be suitable for grey balance color cast correction method, difference is the generating mode of luminance graph.When single channel is impaired, brightness calculation should based on other two non-marred channel, suppose that channel B is impaired, then brightness should be: when L=0.4*Lr+0.4*Lg+0.2*Lb binary channels is impaired, brightness calculation should based on non-marred channel, suppose that G, channel B are impaired, then brightness should be: L=0.6*Lr+0.2*Lg+0.2*Lb.
Embodiment one:
Be example if Fig. 4 is channel offset type colour cast, its integral color is partially red, method of the present invention is adopted first Fig. 4 to be identified as channel offset type, then for carrying out process, its result is as shown in Fig. 4 (a), and its color true nature, with the processing mode that Fig. 4 (b) is traditional, its color is still partially red, and the effect of process of the present invention is better.
Embodiment two:
Be example if Fig. 5 is single channel type of damaged colour cast, the photo that its entirety was taken under street lamp at night, entirety is partially yellow, adopt method of the present invention first Fig. 5 to be identified as single channel type of damaged, then for carrying out process, its result is as shown in Fig. 5 (a), its color true nature, with the processing mode that Fig. 5 (b) is traditional, whiten in its integral color basket, the effect of process of the present invention is better.
Embodiment three:
Be example if Fig. 6 is binary channels type of damaged colour cast, its overall kermesinus partially, method of the present invention is adopted first Fig. 6 to be identified as binary channels type of damaged, then for carrying out process, its result as shown in Fig. 6 (a), its color true nature, with the processing mode that Fig. 6 (b) is traditional, its integral color yellowing, and still with redness, the effect of process of the present invention is better.
The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, without departing from the inventive concept of the premise; can also make some improvements and modifications, these improvements and modifications also should be considered as in scope.
Claims (7)
1., based on a colour cast classification processing method for histogram feature, it is characterized in that, comprise the steps:
1.1) channel information amount is calculated:
1.2) according to channel information amount, image is divided into single channel type of damaged, binary channels type of damaged, channel offset type three class;
1.3) select corresponding respectively according to single channel type of damaged, binary channels type of damaged, channel offset type three class
Pattern colour bias correcting method.
2. a kind of colour cast classification processing method based on histogram feature according to claim 1, it is characterized in that, calculate channel information amount and calculate road histogram width for amounting to, the method that histogram width calculates refers to statistical pixel number from histogram both sides respectively, until reach total pixel 5%, then this place is distributing edge, and border, both sides difference is histogram distribution width.
3. a kind of colour cast classification processing method based on histogram feature according to claim 2, it is characterized in that, described single channel type of damaged is under rgb space, a certain channel histogram width fewer than other two passage 100 time, be considered as single channel impaired.
4. a kind of colour cast classification processing method based on histogram feature according to claim 3, is characterized in that, described binary channels is impaired is under rgb space, a certain channel histogram width more than other two passage 100 time, be considered as binary channels impaired.
5. a kind of colour cast classification processing method based on histogram feature according to claim 4, it is characterized in that, described channel offset type is under rgb space, and histogram width difference is less than 100.
6. a kind of colour cast classification processing method based on histogram feature according to claim 5, it is characterized in that, single channel type of damaged, binary channels type of damaged all adopt grey balance color cast correction method, adopt the generating mode of different luminance graphs according to single channel type of damaged, binary channels type of damaged.
7. a kind of colour cast classification processing method based on histogram feature according to claim 6, is characterized in that, when carrying out color cast correction for channel offset type employing subchannel matlab method, RETINEX bearing calibration.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510305816.XA CN105096265B (en) | 2015-06-04 | 2015-06-04 | A kind of colour cast classification processing method based on histogram feature |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510305816.XA CN105096265B (en) | 2015-06-04 | 2015-06-04 | A kind of colour cast classification processing method based on histogram feature |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105096265A true CN105096265A (en) | 2015-11-25 |
CN105096265B CN105096265B (en) | 2018-09-21 |
Family
ID=54576616
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510305816.XA Active CN105096265B (en) | 2015-06-04 | 2015-06-04 | A kind of colour cast classification processing method based on histogram feature |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105096265B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101383A (en) * | 2020-09-15 | 2020-12-18 | 遵义师范学院 | Color cast image identification method |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102129674A (en) * | 2010-12-17 | 2011-07-20 | 北京优纳科技有限公司 | Self-adaptation color balance correction method for color image |
US20140098098A1 (en) * | 2012-10-09 | 2014-04-10 | E-On Software | Method for accelerating monte carlo renders |
CN103854261A (en) * | 2014-02-25 | 2014-06-11 | 上海理工大学 | Method for correcting color cast images |
CN104202596A (en) * | 2014-09-17 | 2014-12-10 | 西安电子科技大学 | Image color-cast detection method and system applied to intelligent terminal |
-
2015
- 2015-06-04 CN CN201510305816.XA patent/CN105096265B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102129674A (en) * | 2010-12-17 | 2011-07-20 | 北京优纳科技有限公司 | Self-adaptation color balance correction method for color image |
US20140098098A1 (en) * | 2012-10-09 | 2014-04-10 | E-On Software | Method for accelerating monte carlo renders |
CN103854261A (en) * | 2014-02-25 | 2014-06-11 | 上海理工大学 | Method for correcting color cast images |
CN104202596A (en) * | 2014-09-17 | 2014-12-10 | 西安电子科技大学 | Image color-cast detection method and system applied to intelligent terminal |
Non-Patent Citations (1)
Title |
---|
江志: "基于Lab色度空间的色偏检测技术研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101383A (en) * | 2020-09-15 | 2020-12-18 | 遵义师范学院 | Color cast image identification method |
CN112101383B (en) * | 2020-09-15 | 2022-10-21 | 遵义师范学院 | Color cast image identification method |
Also Published As
Publication number | Publication date |
---|---|
CN105096265B (en) | 2018-09-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9619873B2 (en) | Image processing device and image processing method | |
CN107507142B (en) | Signal lamp image processing method | |
CN103065334A (en) | Color cast detection and correction method and device based on HSV (Hue, Saturation, Value) color space | |
CN102883168A (en) | White balance processing method directed towards atypical-feature image | |
CN102737221B (en) | Method and apparatus for vehicle color identification | |
CN102819728A (en) | Traffic sign detection method based on classification template matching | |
CN107730481A (en) | A kind of traffic lights image processing method and traffic lights image processing apparatus | |
CN104778833B (en) | The method for identifying traffic lights | |
CN103491357A (en) | Auto white balance treatment method of image sensor | |
CN105848354B (en) | A kind of shading process being used in intelligent illuminating system and system | |
Zakir et al. | Road sign segmentation based on colour spaces: A Comparative Study | |
CN113748426B (en) | Content aware PQ range analyzer and tone mapping in real-time feed | |
CN105323495A (en) | Region based shutter adaptation method for image exposure | |
CN102509095A (en) | Number plate image preprocessing method | |
CN104766276B (en) | A kind of color cast correction based on color space | |
CN104915975A (en) | Image processing method and system for simulating crayon colored drawing | |
CN104021527A (en) | Rain and snow removal method in image | |
CN106815587A (en) | Image processing method and device | |
Tai et al. | Automatic white balance algorithm through the average equalization and threshold | |
CN102306276A (en) | Method for identifying color of vehicle body in video vehicle image based on block clustering | |
CN104574377A (en) | Method for correcting yellow cast of red lamp of electronic police | |
CN103500457B (en) | A kind of method of video image color cast detection | |
CN105096265A (en) | Colour offset classification method based on histogram features | |
CN102768758B (en) | Improved color image unsharp masking (USM) method | |
CN103679658B (en) | A kind of image processing method according to the decolouring of dominant hue intelligence |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |