CN107886482A - Improve the method and device of Bayer picture contrasts - Google Patents
Improve the method and device of Bayer picture contrasts Download PDFInfo
- Publication number
- CN107886482A CN107886482A CN201711096682.0A CN201711096682A CN107886482A CN 107886482 A CN107886482 A CN 107886482A CN 201711096682 A CN201711096682 A CN 201711096682A CN 107886482 A CN107886482 A CN 107886482A
- Authority
- CN
- China
- Prior art keywords
- view data
- dark primary
- bayer
- primary view
- module
- 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 41
- 238000001914 filtration Methods 0.000 claims abstract description 31
- 230000002146 bilateral effect Effects 0.000 claims abstract description 24
- 238000007726 management method Methods 0.000 claims description 27
- 230000010354 integration Effects 0.000 claims description 24
- 238000003860 storage Methods 0.000 claims description 20
- 238000009825 accumulation Methods 0.000 claims description 12
- 238000004458 analytical method Methods 0.000 claims description 10
- 230000003139 buffering effect Effects 0.000 claims description 9
- 230000006872 improvement Effects 0.000 claims description 9
- 230000005540 biological transmission Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000007619 statistical method Methods 0.000 claims description 4
- 238000006073 displacement reaction Methods 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 9
- 230000015654 memory Effects 0.000 description 21
- 230000003068 static effect Effects 0.000 description 20
- 238000010586 diagram Methods 0.000 description 8
- 239000003086 colorant Substances 0.000 description 3
- 238000004088 simulation Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000004297 night vision Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000011946 reduction process Methods 0.000 description 1
Classifications
-
- G06T5/92—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/20—Image enhancement or restoration by the use of local operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration by the use of histogram techniques
-
- 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
- G06T2207/20024—Filtering details
- G06T2207/20028—Bilateral filtering
Abstract
The present invention provides the method and device for improving Bayer picture contrasts, including：The dark primary view data of the bottom of pixel cell, middle part and top in current Bayer view data is obtained respectively；By the dark primary view data timesharing management of acquisition, Zuo Ding regions TL and You Ding regions TR integrated value are obtained by the dark primary view data at top, left bottom region BL and right bottom region BR integrated value are obtained by the dark primary view data of bottom, and form integrogram；By integrogram, Zuo Ding regions TL, You Ding region TR, left bottom region BL and right bottom region BR integrated value after integrated treatment, bilateral filtering processing is carried out, obtains filtered dark primary view data；Filtered dark primary view data calculating is handled, obtains the Bayer view data after contrast lifting.By the above-mentioned means, while the contrast and detailed information of image is ensured, simplify intermediate treatment process, reduce the delay of processing.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to the method and device for improving Bayer picture contrasts.
Background technology
During image taking and video acquisition, when being frequently encountered night vision or poor backlight, now obtain
Picture contrast is low, and detailed information obscures, and inconvenience is brought to successive image analysis work.Can be with using image enhancement technique
The effectively performance of system in low light situations such as lifting video acquisition, image recognition, Dynamic Recognition.
For coloured image, it is necessary to gather a variety of most basic colors, such as tri- kinds of colors of RGB, simplest method is exactly
With the method for filter, red filter is through red wavelength, and green filter is through the wavelength of green, blue filter transmission
The wavelength of blueness.If gathering tri- Essential colour of RGB, three pieces of filters are needed, it is so expensive, and bad manufacture, because
It all must assure that each pixel aligns for three pieces of filters.When with bayer forms, solve this well and ask
Topic.Bayer images are the different colors set on one piece of filter, are found by analyzing perception of the human eye to color.Bayer
Pattern is color mode, is widely used in CCD and CMOS camera.The contrast of Bayer view data is directed in the prior art
Degree method for improving mainly comprises the following steps：Interpolation processing first is carried out for Bayer view data, obtains the figures of the RGB after interpolation
As data, then its dark primary view data is obtained based on rgb image data；Its integrogram is calculated for dark primary image, is led to
Cross integrogram and bilateral filtering processing is done to the dark primary image obtained, obtain filtered dark primary view data；For RGB
View data, filtered dark primary view data carry out statistics with histogram, obtain R/G/B (tri- components of pixel R, G, B
Value), dark primary view data histogram；Cumulative statistics is carried out for filtered dark primary view data simultaneously, is filtered
Dark primary view data summation in window；For R/G/B, dark primary view data histogram, dark primary image in filter window
Data summation carries out comprehensive analysis, obtains yield value, dark average and air light value corresponding to R/G/B；Updated using every frame
Dark primary view data after air light value, dark average and filtering corresponding to obtained yield value and R/G/B, to original
Bayer view data carries out calculating processing, finally gives the Bayer view data after contrast lifting.
In the prior art, it is necessary to enter for Bayer view data, the front/rear dark primary view data of bilateral filtering, integrogram
Row caching process, it is necessary to repeatedly access internal memory while committed memory；In addition, in former method, except needing statistics/safeguard R/
G/B histograms, it is also necessary to do the calculating processing of multiple pilot process.Therefore, the existing contrast for Bayer view data
In method for improving, graphics process or Digital Signal Processing are realized and all meet very much requirement of real time, or even processing delay occur
Reach several seconds levels.
Therefore, it is necessary to which a kind of improve Bayer picture contrast schemes, the contrast and detailed information of image can ensured
While, improve the delay of processing procedure.
The content of the invention
Present invention solves the technical problem that being to provide the method and device for improving Bayer picture contrasts, ensureing image
Contrast and detailed information while, simplify intermediate treatment process, reduce the delay of processing.
In order to solve the above technical problems, the invention provides the method for improving Bayer picture contrasts, including：Improve
The method of Bayer picture contrasts, including bottom, middle part and the top of pixel cell in current Bayer view data are obtained respectively
The dark primary view data in portion；By the dark primary view data timesharing management of acquisition, pass through the dark primary picture number at the top
According to Zuo Ding regions TL and You Ding regions TR integrated value is obtained, left bottom region is obtained by the dark primary view data of the bottom
BL and right bottom region BR integrated value, and form integrogram；Pass through the integrogram, the Zuo Ding regions TL, the You Ding areas
Domain TR, the left bottom region BL and the right bottom region BR integrated value obtain dark primary figure in filter window after integrated treatment
As data summation, dark primary view data summation in the filter window is subjected to bilateral filtering processing, obtained filtered dark
Primary color image data；The filtered dark primary view data calculating is handled, obtains the Bayer images after contrast lifting
Data.
Wherein, the integrogram is realized by the way of integration row.
Further, it is described to pass through the integrogram, the Zuo Ding regions TL, the You Ding regions TR, the Zuo Di areas
Domain BL and the right bottom region BR integrated value obtain the step of dark primary view data summation in filter window after integrated treatment
Suddenly specifically include：First merge the Zuo Ding regions TL and the right bottom region BR integrated value, then subtract the You Ding regions TR
With the integrated value of the left bottom region BL, dark primary view data summation in filter window is obtained.
Further, it is described to obtain the dark of the bottom of pixel cell in current Bayer view data, middle part and top respectively
The specific steps of primary color image data include：Current Bayer view data is subjected to interpolation processing, obtains the figures of the RGB after interpolation
As data；According to the dark primary of the rgb image data after the interpolation, the respectively bottom of acquisition pixel cell, middle part and top
View data.
Further, it is described to handle the filtered dark primary view data calculating, after obtaining contrast lifting
The specific steps of Bayer view data include：Carried out according to the rgb image data and the filtered dark primary image straight
Side's figure statistics, obtain R/G/B view data accumulation histogram, dark primary view data histogram, the accumulation of dark primary view data
Total value；By the R/G/B view data accumulation histogram, the dark primary view data histogram, the dark primary picture number
Comprehensive analysis is carried out according to cusum, obtains air light value corresponding to yield value and R/G/B view data；Utilize Bayer images
It is air light value corresponding to the yield value and R/G/B view data that data obtain and the filtered dark primary image, described
Bayer view data, COMPREHENSIVE CALCULATING handle to obtain the Bayer view data after contrast lifting.
To solve the above problems, the present invention also provides the device for improving Bayer picture contrasts, including：Storage manager,
The storage manager is used for the caching for realizing the Bayer picturedeep evidences in bilateral filtering window；Module is run in bottom, described
Bottom operation module is used for the bottom dark primary view data for obtaining pixel cell in Bayer view data, and according to the bottom
Left bottom region BL and right bottom region BR integrated value is calculated in portion's dark primary view data；Run module, the top in top
Operation module is used for the top dark primary view data for obtaining pixel cell in Bayer view data, and dark according to the top
Zuo Ding regions TL and You Ding regions TR integrated value is calculated in primary color image data；Run module, the middle part operation in middle part
Module is used to form integrogram, passes through the integrogram, the Zuo Ding regions TL, the You Ding regions TR, the left bottom region
BL and the right bottom region BR integrated value obtain dark primary view data summation in filter window after integrated treatment, by described in
Dark primary view data summation carries out bilateral filtering processing in filter window, obtains filtered dark primary view data, uses
The filtered dark primary view data tool carries out calculating processing, obtains the Bayer view data after contrast lifting.
Wherein, the bottom operation module includes：Row buffering management module；For resetting the Bayer picture numbers before interpolation
According to color filter array interpolation module；Dark primary computing module, it is dark primary picture number by choosing the minimum value in RGB image
According to；Row management module is integrated, for realizing the management of the integration row in filter window corresponding to top and bottom；The top fortune
Row module is identical with bottom operation modular structure.
Wherein, the middle part operation module includes：Row buffering management module, for resetting the Bayer picture numbers before interpolation
According to color filter array interpolation module；Color filter array interpolation module, for Bayer view data interpolation calculations, obtain RGB
Image；Dark primary computing module, it is dark primary view data by choosing the minimum value in RGB image；Barrel shifter shifts, it is used for
Specify direction and the displacement of Bayer view data transmission；Bilateral filtering computing module, for realizing to dark primary view data
Filtering and smoothing processing, obtain filtered dark primary image；Statistical analysis computing module, the computing module are used for total score
Analysis handles the dark primary view data, the filtered dark primary view data, after contrast lifting is obtained after calculating
Bayer view data.
Wherein, the bottom operation module, top operation module and middle part operation module are parallel modules.
Wherein, the storage manager includes：Storage control, the storage control is in bilateral filtering window
The caching of Bayer picturedeep evidences；Arbitration manager, the arbitration manager are responsible for the Bayer picturedeep evidences of multiple directions
Read-write management.
Wherein, the integration row management module is updated and storage simulation dark primary using 2 SRAM (static memory)
The integrated value of picturedeep evidence.
The beneficial effects of the invention are as follows：Prior art is different from, the present invention proposes to improve the method for Bayer picture contrasts
And device, methods and apparatus of the present invention are ensureing the contrast and details of image using streamline, parallel data processing structure
While information, simplify intermediate treatment process, reduce the delay of processing, the perfect processing procedure of Bayer images.Further
, the integrogram needed for bilateral filtering is converted to integration line mode and is managed by methods and apparatus of the present invention, optimizes dark-state
Image, suppress overly bright region, complete the lifting of the overall contrast of image.
Brief description of the drawings
Fig. 1 is the schematic flow sheet for the embodiment of method one that the present invention improves Bayer picture contrasts；
Fig. 2 is each domain integral schematic diagram of dark primary view data of the present invention；
Fig. 3 is dark primary view data integration row schematic diagram of the present invention；
Fig. 4 is that dark primary view data summation calculates principle schematic in filter window of the present invention；
Fig. 5 is the apparatus structure schematic diagram that first embodiment of the invention improves Bayer picture contrasts；
Fig. 6 is the apparatus structure schematic diagram that second embodiment of the invention improves Bayer picture contrasts；
Fig. 7 is the apparatus structure schematic diagram that third embodiment of the invention improves Bayer picture contrasts.
Embodiment
Referring to Fig. 1, Fig. 1 is the schematic flow sheet for the embodiment of method one that the present invention improves Bayer picture contrasts,
The method of the improvement Bayer picture contrasts of the present embodiment comprises the following steps：
101：The dark primary image of the bottom of pixel cell, middle part and top in current Bayer view data is obtained respectively
Data.
Specifically, current Bayer view data is carried out into interpolation processing, the rgb image data after interpolation is obtained；Further according to
Rgb image data after the interpolation, obtain the dark primary view data at the bottom of pixel cell, middle part and top.
In the present embodiment, by choosing the minimum value in RGB image come as dark primary view data, bottom operation module
The bottom dark primary view data of pixel cell in Bayer view data is obtained, top operation module obtains Bayer view data
The top dark primary view data of middle pixel cell, middle part operation module obtain the dark primary figure of part in Bayer view data
As data.
102：By dark primary view data timesharing management, by the dark primary view data at top obtain Zuo Ding regions TL and
You Ding regions TR integrated value, left bottom region BL and left bottom region BL integration are obtained by the dark primary view data of bottom
Value, and form integrogram.
Preferably, in order to synchronously be integrated, module is run at top and bottom operation module carries out timesharing management, bottom
Operation module directly obtains Bayer view data, and top operation module then obtains the Bayer view data in buffer storage.
As shown in Fig. 2 wherein, Zuo Ding regions TL201, You Ding region TR202, left bottom region BL203 and left bottom region
BL204, respective pixel cell is represented respectively to the region of basic point 205.
In the present embodiment, integrogram is realized by the way of integration row.In a specific embodiment, as shown in figure 3,
Using integration row come simulated implementation integrogram function, according to calculated in filter window accumulation and required Zuo Ding regions TL301 and
You Ding regions TR302 and left bottom region BL303 and right bottom region BR304 integrated values, module is run by top and bottom is transported
Row module branch, timesharing are managed；Every group of integration row uses 2 static memories with time-sharing format come the pipe of analog integration value
The corresponding top integration row of reason, wherein static memory A and B；And static memory C and D correspond to the integration row of bottom；Work as Bayer
Before view data the first row data reach, the zeros data in static memory A, B, C and D；The line number of Bayer view data first
According to reach when, static memory A will using static memory B for reference, from static memory B at the same read (i-1, j-1),
The integrated value of (i-1, j)；The value of (i, j-1) is read from static memory A；After being calculated by integrated value formula, by integrated value
(i, j) is updated into static memory A correspondence positions；When the second row of Bayer view data data reach, now static storage
The integration data of the first row data has been preserved in device A；Static memory B will calculate current line using static memory A as reference
Integrated value, and updated in static memory B；When picture frame the third line data reach, now static memory A, B be again
Secondary switching, static memory A calculates the integration diagram data of current line again using static memory B as reference, and updates to static state
Memory A；Static memory C and D way to manage are identical with static memory A and B.By that analogy, until terminating by calculating
Obtain integrogram.
103：Passed through by integrogram, Zuo Ding regions TL, You Ding region TR, left bottom region BL and right bottom region BR integrated value
Dark primary view data summation in filter window is obtained after integrated treatment, dark primary view data summation in filter window is carried out
Bilateral filtering processing, obtains filtered dark primary view data.
Preferably, in filter window the Computing Principle of dark primary view data summation as shown in figure 4, Integral (i, j)
Information to be stored in integrogram is integration of (i, the j) pixel cell to all pixels sum of basic point (0,0) (405 in figure)
Value, Pixel (i, j) is pixel cell integrated value.
Pixel (0,0), the integrated value of its integrogram correspondence position are：
Integral (0,0)=Pixel (0,0).
Pixel (0,1), the integrated value of its integrogram correspondence position are：
Integral (0,1)=Pixel (0,1)+Integral (0,0).
Pixel (1,0), the integrated value of its integrogram correspondence position are：
Integral (1,0)=Pixel (1,0)+Integral (0,0).
Pixel (1,1), the integrated value of its integrogram correspondence position are：
Integral (1,1)=Integral (1,0)+Integral (0,1)-Integral (0,0)+Pixel (1,1).
Then for Pixel (i, j), as shown in Fig. 4 401~404, the integrated value of its integrogram correspondence position is：
Integral (i, j)=Integral (i, j-1)+Integral (i-1, j)-Integral (i-1, j-1)+
Pixel(i,j)。
By the above results, dark primary view data summation is by merging Zuo Ding regions TL and right bottom region in filter window
BR integrated value, the integration for subtracting You Ding regions TR and left bottom region BL are worth to, dark primary view data summation meter in window
Calculate formula：
Sum=BR+TL-TR-BL (1)
(wherein, Sum, dark primary view data summation in window；BR, the integrated value in right bottom region；The integration in TL, Zuo Ding region
Value；The integrated value in TR, You Ding region；BL, the integrated value in left bottom region)
In a specific embodiment, filter window radius is R, and the size of whole filter window is (2R+1) * (2R+
1), now module, the distance of two lateral extents homogeneous poor (2R+1), Jiang Zuoding areas are run in the bottom operation module in window and top
Domain TL, You Ding region TR, left bottom region BL and right bottom region BR integrated value import above-mentioned formula (1), obtain whole bilateral filter
Dark primary view data summation in ripple device (2R+1) * (2R+1) window, dark primary view data summation in obtained window is entered
The processing of row bilateral filtering, realizes the noise reduction process to Bayer view data.
104：Filtered dark primary view data calculating is handled, obtains the Bayer view data after contrast lifting.
Specifically, carrying out statistics with histogram according to rgb image data and filtered dark primary image first, R/G/ is obtained
B view data accumulation histogram, dark primary view data histogram, dark primary view data cusum；Again by R/G/B images
Data accumulation histogram, dark primary view data histogram, dark primary view data cusum carry out comprehensive analysis, are increased
Air light value corresponding to benefit value and R/G/B view data；Yield value and the R/G/B figure finally obtained using Bayer view data
Air light value and filtered dark primary image, Bayer view data as corresponding to data, COMPREHENSIVE CALCULATING, which is handled, to be contrasted
Bayer view data after degree lifting.
Be different from prior art, the present embodiment by bottom, middle part and top run module obtain respectively bottom, middle part and
The dark primary view data at top, then by the domain integral of bottom and top to obtain dark primary view data in filter window total
With COMPREHENSIVE CALCULATING processing, so as to obtain the Bayer view data after contrast lifting.By the above-mentioned means, ensureing image
While contrast and detailed information, simplify intermediate treatment process, reduce the delay of processing, the perfect processing of Bayer images
Process.
Refering to Fig. 5, Fig. 5 is the apparatus structure schematic diagram that first embodiment of the invention improves Bayer picture contrasts, mainly
Including：Storage manager 1, for realizing the caching of the Bayer picturedeep evidences in bilateral filtering window, storage manager 1 is pressed
Include successively according to transmission direction：Storage control 18, arbitration manager 19；Bottom operation module (not marked in figure), for obtaining
The bottom dark primary view data of pixel cell in Bayer view data is taken, and according to the bottom dark primary view data meter
Calculation obtains left bottom region BL and right bottom region BR integrated value, and bottom operation module includes successively according to transmission direction：Row buffering
Management module 2, color filter array interpolation module 3, dark primary computing module 4 and integration row management module 5；Run module in top
(not marked in figure), for obtaining the top dark primary view data of pixel cell in Bayer view data, and according to the top
Zuo Ding regions TL and You Ding regions TR integrated value is calculated in portion's dark primary view data, and top operation module is run with bottom
Modular structure is identical, and top operation module includes successively according to transmission direction：Row buffering management module 14, color filter array are inserted
It is worth module 15, dark primary computing module 16 and integration row management module 17；Middle part operation module (not marked in figure), for obtaining
Dark primary view data in filter window corresponding to middle part corresponding to Bayer image lines；Simultaneously to Bayer image lines and dark original
Color image row carries out registration process；Middle part operation module is used to receive the left bottom region BL integrated values, right bottom region BR product
Score value, Zuo Ding regions TL integrated values and You Ding regions TR integrated values, comprehensive tetra- values of BL, BR, TL, TR, are obtained in filter window
Dark primary view data summation, then after bilateral filtering calculating processing, obtain filtered dark primary view data；Run at middle part
Module includes：It is row buffering management module 6, color filter array interpolation module 7, dark primary computing module 8, barrel shifter shifts 9, double
Side filters computing module 10 and statistical analysis computing module.Wherein, statistical analysis computing module includes statistical module 11, analyzes mould
Block 12, computing module 13.Bottom operation module, top operation module and middle part operation module are parallel modules, and using stream
Line structure.
In the present embodiment, row buffering management module (2,6,14 in figure) is used to reset the Bayer view data before interpolation；Color
Color filter array interpolating module (3,7,15 in figure) is used for Bayer view data interpolation calculations, obtains RGB image；Dark primary meter
It is dark primary view data that module (4,8,16 in figure), which is calculated, by choosing the minimum value in RGB image；Integrate row management module (figure
In 5,17) be used to realize in filter window the management of integration row corresponding to top and bottom, and using 2 SRAM updating and
The integrated value of storage simulation dark primary picturedeep evidence；Barrel shifter shifts 9 are used for the direction for specifying Bayer view data to transmit
With displacement；Bilateral filtering computing module 10 is used to realize the filtering to dark primary view data and smoothing processing, after obtaining filtering
Dark primary image；Statistical module 11 is used to filtered dark primary image carrying out statistics with histogram, obtains R/G/B picture numbers
According to accumulation histogram, dark primary view data histogram, dark primary view data cusum；Analysis module 12 schemes R/G/B
As data accumulation histogram, dark primary view data histogram, dark primary view data cusum progress comprehensive analysis, obtain
Air light value corresponding to yield value and R/G/B view data；Yield value that computing module 13 obtains according to Bayer view data and
Air light value corresponding to R/G/B view data and filtered dark primary image, Bayer view data, COMPREHENSIVE CALCULATING processing
Obtain the Bayer view data after contrast lifting；Storage control 18 is used for the Bayer image lines in bilateral filtering window
The caching of data；Arbitration manager 19 is responsible for the read-write management of the Bayer picturedeep evidences of multiple directions.
In a specific embodiment, Bayer view data inputs bottom operation module and storage manager 1 respectively,
By storage control 18 and are sent into top operation module by view data timesharing for arbitration manager 19 and module is run at middle part；
The dark primary view data of the bottom of pixel cell, middle part and top is successively by row buffering management mould in Bayer view data
Block (2,6,14 in figure), color filter array interpolation module (3,7,15 in figure) and dark primary computing module (4,8,16 in figure) enter
After row processing, respective dark primary view data and rgb image data are obtained；The dark primary view data of top and bottom via
Row management module (5,17 in figure) is integrated, the management of the integration row in filter window corresponding to top and bottom is realized, obtains a left side
Region TL, You Ding region TR, left bottom region BL and right bottom region BR integrated value are pushed up, and forms integrogram；Pass through integrogram, a left side
Filtered after the integrated calculating integrated treatment of top region TL, You Ding region TR, left bottom region BL and right bottom region BR integrated value
Dark primary view data summation in ripple window, by dark primary view data summation in filter window via bilateral filtering computing module
After the processing of 10 bilateral filterings, filtered dark primary view data is obtained；By statistical module 11, according to rgb image data and
Filtered dark primary image carries out statistics with histogram, and it is straight to obtain R/G/B view data accumulation histogram, dark primary view data
Fang Tu, dark primary view data cusum；By analysis module 12 by R/G/B view data accumulation histogram, dark primary figure
As data graphs, dark primary view data cusum progress comprehensive analysis, obtain yield value and R/G/B view data is corresponding
Air light value；It is big corresponding to the yield value and R/G/B view data obtained in computing module 13 using Bayer view data
Gas light value and filtered dark primary image, Bayer view data, COMPREHENSIVE CALCULATING are handled after obtaining contrast lifting
Bayer view data.
In one embodiment, referring to Fig. 6, Fig. 6 is the device that second embodiment of the invention improves Bayer picture contrasts
Structural representation, the image data format of input is RGB image, it is not necessary to do color filter array interpolation processing (CFAI,
Color Filter Array Interpolation) processing, the color filter array interpolation that fades that can be gone in above-described embodiment
Module, rgb image data is directly obtained to carry out subsequent treatment, specific apparatus structure is similar to the above embodiments, herein not
Illustrate again.
In another specific embodiment, referring to Fig. 7, Fig. 7, which is third embodiment of the invention, improves Bayer images pair
Than the apparatus structure schematic diagram of degree, the data image of input is YUV image, needs during input to add yuv format and turns to rgb format
Parallel operation 22, delivery outlet need to add rgb format to yuv format converter 23, and specific steps are similar to the above embodiments, herein no longer
Illustrate.
The device of implementation method Bayer picture contrasts lifting also extends to RGB FHD (1920*1080) or UD
The scan picture of resolution factors such as (3840*2160), is not limited thereto.
Prior art is different from, the present invention proposes to improve the method and device of Bayer picture contrasts, method of the invention
With device using streamline, parallel data processing structure, while the contrast and detailed information of image is ensured, simplify middle
Processing procedure, reduces the delay of processing, the perfect processing procedure of Bayer images.Further, methods and apparatus of the present invention
Integrogram needed for bilateral filtering is converted into integration line mode to be managed, optimizes dark-state image, suppresses overly bright region, complete
The lifting of the overall contrast of image.
Embodiments of the present invention are the foregoing is only, are not intended to limit the scope of the invention, it is every to utilize this
The equivalent structure or equivalent flow conversion that description of the invention and accompanying drawing content are made, or directly or indirectly it is used in other correlations
Technical field, it is included within the scope of the present invention.
Claims (10)
1. improve the method for Bayer picture contrasts, it is characterised in that including：
The dark primary view data of the bottom of pixel cell, middle part and top in current Bayer view data is obtained respectively；
By the dark primary view data timesharing management of acquisition, Zuo Ding regions TL is obtained by the dark primary view data at the top
With You Ding regions TR integrated value, obtain left bottom region BL's and right bottom region BR by the dark primary view data of the bottom
Integrated value, and form integrogram；
Pass through the integrogram, the Zuo Ding regions TL, the You Ding regions TR, the left bottom region BL and the right bottom region
BR integrated value obtains dark primary view data summation in filter window after integrated treatment, by dark primary in the filter window
View data summation carries out bilateral filtering processing, obtains filtered dark primary view data；
The filtered dark primary view data calculating is handled, obtains the Bayer view data after contrast lifting.
2. the method for improvement Bayer picture contrasts according to claim 1, it is characterised in that the integrogram uses
The mode of integration row is realized.
3. the method for improvement Bayer picture contrasts according to claim 1, it is characterised in that described to pass through the product
Component, the Zuo Ding regions TL, the You Ding regions TR, the left bottom region BL and the right bottom region BR integrated value are through comprehensive
The step of obtaining dark primary view data summation in filter window after conjunction processing specifically includes：
First merge the Zuo Ding regions TL and the right bottom region BR integrated value, then subtract the You Ding regions TR and the left side
Bottom region BL integrated value, obtain dark primary view data summation in filter window.
4. the method for improvement Bayer picture contrasts according to claim 1, it is characterised in that described obtain respectively is worked as
The specific steps of the dark primary view data of the bottom of pixel cell, middle part and top include in preceding Bayer view data：
Current Bayer view data is subjected to interpolation processing, obtains the rgb image data after interpolation；
According to the dark primary image of the rgb image data after the interpolation, the respectively bottom of acquisition pixel cell, middle part and top
Data.
5. the method for improvement Bayer picture contrasts according to claim 1, it is characterised in that described by the filtering
Dark primary view data calculating processing afterwards, obtaining the specific steps of the Bayer view data after contrast lifting includes：
Statistics with histogram is carried out according to the rgb image data and the filtered dark primary image, obtains R/G/B picture numbers
According to accumulation histogram, dark primary view data histogram, dark primary view data cusum；
By the R/G/B view data accumulation histogram, the dark primary view data histogram, the dark primary view data
Cusum carries out comprehensive analysis, obtains air light value corresponding to yield value and R/G/B view data；
After air light value corresponding to the yield value and R/G/B view data obtained using Bayer view data and the filtering
Dark primary image, the Bayer view data, COMPREHENSIVE CALCULATING handles to obtain the Bayer view data after contrast lifting.
6. improve the device of Bayer picture contrasts, it is characterised in that including：
Storage manager, the storage manager are used for the caching for realizing the Bayer picturedeep evidences in bilateral filtering window；
Module is run in bottom, and the bottom operation module is used for the bottom dark primary for obtaining pixel cell in Bayer view data
View data, and left bottom region BL and right bottom region BR integrated value is calculated according to the bottom dark primary view data；
Module is run at top, and the top operation module is used for the top dark primary for obtaining pixel cell in Bayer view data
View data, and Zuo Ding regions TL and You Ding regions TR integrated value is calculated according to the top dark primary view data；
Module is run at middle part, and the middle part operation module is used to form integrogram, passes through the integrogram, the Zuo Ding regions
TL, the You Ding regions TR, the left bottom region BL and the right bottom region BR integrated value are filtered after integrated treatment
Dark primary view data summation in window, dark primary view data summation in the filter window is subjected to bilateral filtering processing,
Filtered dark primary view data is obtained, calculating processing is carried out using the filtered dark primary view data tool, obtains
Bayer view data after contrast lifting.
7. the device of improvement Bayer picture contrasts according to claim 6, it is characterised in that：Run mould in the bottom
Block includes：
Row buffering management module；For resetting the Bayer image data color filter array interpolating modules before interpolation；
Dark primary computing module, it is dark primary view data by choosing the minimum value in RGB image；
Row management module is integrated, for realizing the management of the integration row in filter window corresponding to top and bottom；
The top operation module is identical with bottom operation modular structure.
8. the device of improvement Bayer picture contrasts according to claim 6, it is characterised in that：Run mould in the middle part
Block includes：
Row buffering management module, for resetting the Bayer image data color filter array interpolating modules before interpolation；
Color filter array interpolation module, for Bayer view data interpolation calculations, obtain RGB image；
Dark primary computing module, it is dark primary view data by choosing the minimum value in RGB image；
Barrel shifter shifts, for specifying direction and the displacement of the transmission of Bayer view data；
Bilateral filtering computing module, for realizing filtering and smoothing processing to dark primary view data, obtain filtered dark
Primary colour image；
Statistical analysis computing module, the computing module are used for dark primary view data, the filtering described in comprehensive analysis processing
Dark primary view data afterwards, the Bayer view data after contrast lifting is obtained after calculating.
9. the device of improvement Bayer picture contrasts according to claim 6, it is characterised in that run mould in the bottom
Block, top operation module and middle part operation module are parallel modules.
10. the device of improvement Bayer picture contrasts according to claim 6, it is characterised in that the storage manager
Including：
Storage control, caching of the storage control to the Bayer picturedeep evidences in bilateral filtering window；
Arbitration manager, the arbitration manager are responsible for the read-write management of the Bayer picturedeep evidences of multiple directions.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711096682.0A CN107886482B (en) | 2017-11-07 | 2017-11-07 | Method and device for improving Bayer image contrast |
| PCT/CN2017/117315 WO2019090909A1 (en) | 2017-11-07 | 2017-12-20 | Method and device for improving contrast of bayer image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711096682.0A CN107886482B (en) | 2017-11-07 | 2017-11-07 | Method and device for improving Bayer image contrast |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107886482A true CN107886482A (en) | 2018-04-06 |
| CN107886482B CN107886482B (en) | 2020-06-05 |
Family
ID=61779582
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711096682.0A Active CN107886482B (en) | 2017-11-07 | 2017-11-07 | Method and device for improving Bayer image contrast |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN107886482B (en) |
| WO (1) | WO2019090909A1 (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102129146A (en) * | 2010-01-12 | 2011-07-20 | 株式会社尼康 | Image pickup device |
| US20120008862A1 (en) * | 2010-07-08 | 2012-01-12 | Mithun Das Gupta | Bi-Affinity Filter: A Bilateral Type Filter for Color Images |
| CN106504281A (en) * | 2016-12-02 | 2017-03-15 | 中国电子科技集团公司第四十四研究所 | The image quality for being applied to cmos image sensor strengthens and filtering method |
| CN106875358A (en) * | 2017-02-09 | 2017-06-20 | 聚龙智瞳科技有限公司 | Image enchancing method and image intensifier device based on Bayer format |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101848343B (en) * | 2009-03-24 | 2013-04-17 | 财团法人工业技术研究院 | Image sensor with integral image output |
| US20120328160A1 (en) * | 2011-06-27 | 2012-12-27 | Office of Research Cooperation Foundation of Yeungnam University | Method for detecting and recognizing objects of an image using haar-like features |
| CN102938136A (en) * | 2012-07-19 | 2013-02-20 | 中国人民解放军国防科学技术大学 | Method for defogging single images based on Bayer formats rapidly |
| US9224362B2 (en) * | 2013-03-14 | 2015-12-29 | Microsoft Technology Licensing, Llc | Monochromatic edge geometry reconstruction through achromatic guidance |
| CN105991937A (en) * | 2015-03-04 | 2016-10-05 | 深圳市朗驰欣创科技有限公司 | Virtual exposure method and device based on Bayer format image |
| CN106657948A (en) * | 2017-01-18 | 2017-05-10 | 聚龙智瞳科技有限公司 | low illumination level Bayer image enhancing method and enhancing device |
-
2017
- 2017-11-07 CN CN201711096682.0A patent/CN107886482B/en active Active
- 2017-12-20 WO PCT/CN2017/117315 patent/WO2019090909A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102129146A (en) * | 2010-01-12 | 2011-07-20 | 株式会社尼康 | Image pickup device |
| US20120008862A1 (en) * | 2010-07-08 | 2012-01-12 | Mithun Das Gupta | Bi-Affinity Filter: A Bilateral Type Filter for Color Images |
| CN106504281A (en) * | 2016-12-02 | 2017-03-15 | 中国电子科技集团公司第四十四研究所 | The image quality for being applied to cmos image sensor strengthens and filtering method |
| CN106875358A (en) * | 2017-02-09 | 2017-06-20 | 聚龙智瞳科技有限公司 | Image enchancing method and image intensifier device based on Bayer format |
Non-Patent Citations (1)
| Title |
|---|
| 娄静涛 等: ""单幅Bayer格式图像的快速去雾方法"", 《国防科技大学学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107886482B (en) | 2020-06-05 |
| WO2019090909A1 (en) | 2019-05-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107123089B (en) | Remote sensing image super-resolution reconstruction method and system based on depth convolution network | |
| CN106296612B (en) | A kind of stagewise monitor video sharpening system and method for image quality evaluation and weather conditions guidance | |
| CN100574369C (en) | Image processing apparatus and image processing method | |
| EP4109392A1 (en) | Image processing method and image processing device | |
| CN103238335B (en) | Image processing apparatus and image processing method | |
| Gao et al. | Sand-dust image restoration based on reversing the blue channel prior | |
| CN104809700B (en) | A kind of low-light (level) video real time enhancing method based on bright passage | |
| CN106791623A (en) | A kind of panoramic video joining method and device | |
| CN110378848B (en) | Image defogging method based on derivative map fusion strategy | |
| CN106530265B (en) | A kind of Adaptive image fusion method based on chromaticity coordinate | |
| CN107507145A (en) | A kind of underwater picture Enhancement Method based on the stretching of different colours spatially adaptive histogram | |
| CN103886565A (en) | Nighttime color image enhancement method based on purpose optimization and histogram equalization | |
| CN103455979A (en) | Low illumination level video image enhancement method | |
| CN110544213A (en) | Image defogging method based on global and local feature fusion | |
| CN103793888B (en) | Image enhancing method based on main colors of reference image | |
| CN103065334A (en) | Color cast detection and correction method and device based on HSV (Hue, Saturation, Value) color space | |
| CN106097279A (en) | A kind of high-dynamics image Enhancement Method | |
| CN102306307B (en) | Positioning method of fixed point noise in color microscopic image sequence | |
| CN107730453A (en) | Picture quality method for improving | |
| CN111415304A (en) | Underwater vision enhancement method and device based on cascade deep network | |
| CN107610675A (en) | A kind of image processing method and device based on dynamic level | |
| CN116681636A (en) | Light infrared and visible light image fusion method based on convolutional neural network | |
| CN104463806B (en) | Height adaptive method for enhancing picture contrast based on data driven technique | |
| CN108711160A (en) | A kind of Target Segmentation method based on HSI enhancement models | |
| CN105427265A (en) | Color image contrast enhancement method and system |
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 |