CN107025633A - A kind of image processing method and device - Google Patents
A kind of image processing method and device Download PDFInfo
- Publication number
- CN107025633A CN107025633A CN201610067209.9A CN201610067209A CN107025633A CN 107025633 A CN107025633 A CN 107025633A CN 201610067209 A CN201610067209 A CN 201610067209A CN 107025633 A CN107025633 A CN 107025633A
- Authority
- CN
- China
- Prior art keywords
- image
- brightness
- area
- profile
- pixel
- 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
- 238000003672 processing method Methods 0.000 title claims abstract description 14
- 238000012545 processing Methods 0.000 claims abstract description 74
- 238000011002 quantification Methods 0.000 claims abstract description 44
- 238000013139 quantization Methods 0.000 claims abstract description 26
- 238000000034 method Methods 0.000 claims abstract description 15
- 235000019557 luminance Nutrition 0.000 claims description 29
- 238000000605 extraction Methods 0.000 claims description 5
- 230000000694 effects Effects 0.000 abstract description 15
- 230000007246 mechanism Effects 0.000 abstract description 4
- 230000001815 facial effect Effects 0.000 description 15
- 238000010586 diagram Methods 0.000 description 10
- 230000008569 process Effects 0.000 description 6
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000000926 separation method Methods 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
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
- G06T5/92—Dynamic range modification of images or parts thereof based on global image properties
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- 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
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
-
- 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
- G06T2207/20172—Image enhancement details
- G06T2207/20208—High dynamic range [HDR] image processing
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Image Analysis (AREA)
Abstract
The present invention provides a kind of image processing method and device, and wherein method includes：Pending image is obtained, the first area in pending image is determined, first area is the region comprising default object or a predeterminable area；Brightness Fuzzy Processing is carried out to pending image, the first image is obtained；The profile of first area in the first image is extracted, and according to the profile of first area, calculates the brightness judgement threshold on profile border, and according to brightness judgement threshold, determine the profile boundary information of all profiles in the first image；The thick quantification treatment of monochrome information is carried out to pending image, the second image is obtained；Using the profile boundary information of all profiles in the first image, the profile border in the second image is carried out after intensive treatment, the image after being handled.The filter that the embodiment of the present invention improves image by using contours extract mechanism, increase luminance quantization level mechanism is taken pictures effect, effectively solve conventional images filter shooting effect pit it is excessive the problem of.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to a kind of image processing method and device.
Background technology
Intelligent terminal camera applications are the function of a use widely.In order to lift the practicality of camera applications
Property, many image processing functions are implanted into application, filter is exactly a kind of.But filter of the prior art
Mirror is limited to the simplicity that it designs device, and often treatment effect compares not as people's will.
The content of the invention
The purpose of the embodiment of the present invention is to provide a kind of image processing method and device, it is intended to solve existing skill
Because filter design is simple during intelligent terminal camera applications in art, the obtained effect image caused is unsharp
Problem.
The embodiment of the present invention provides a kind of image processing method, including：
Pending image is obtained, the first area in the pending image is determined, the first area is bag
Region or a predeterminable area containing default object；
Brightness Fuzzy Processing is carried out to the pending image, the first image is obtained；
The profile of first area described in described first image is extracted, and according to the profile of the first area,
The brightness judgement threshold on profile border is calculated, and according to the brightness judgement threshold, determines first figure
The profile boundary information of all profiles as in；
The thick quantification treatment of monochrome information is carried out to the pending image, the second image is obtained；
Using the profile boundary information of all profiles in described first image, to the profile in second image
Border is carried out after intensive treatment, the image after being handled.
Wherein, the first area determined in the pending image, including：
Judge in the pending image with the presence or absence of default object；
When there is default object in the pending image, it is determined that region shared by default object is described the
One region；
When default object is not present in the pending image, it is the first area to determine predeterminable area.
Wherein, it is described when there is default object in the pending image, it is determined that the area shared by default object
Domain is the first area, including：
When there is multiple default objects in the pending image, the maximum default object of area is chosen, and
It is the first area to determine the region shared by the maximum default object of area；
When presetting object in the presence of two size identicals in the pending image, choose and sat with center
The default object of subject distance closely, and determine that the region shared by the default object near with centre coordinate distance is described
First area.
Wherein, the profile for extracting first area described in described first image, and according to described first
The profile in region, calculates the brightness judgement threshold on profile border, including：
The wire-frame image vegetarian refreshments of the first area is extracted, the brightness of the wire-frame image vegetarian refreshments of the first area is counted
Maximum, brightness minimum value or average brightness；
Multiplied according to product, the brightness minimum value of brightness maxima and the first proportionality coefficient with the second proportionality coefficient
Product or average brightness determine the brightness judgement threshold on profile border, wherein first proportionality coefficient is small
1 positive number, second proportionality coefficient is the positive number more than 1.
Wherein, the thick quantification treatment that monochrome information is carried out to the pending image, obtains the second image,
Including：
The pixel of the pending image is extracted, to the pixel of the pending image according to equation below
The grade processing of increase luminance quantization is carried out, second image is obtained：
Y1=[y/Q] * Q
Wherein：Y1 represents the pixel brightness after increasing quantification gradation processing, and y represents original pixel and lighted
Degree, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, and [y/Q] is represented to y
Floor operation is done with Q business.
Wherein, in the utilization described first image all profiles profile boundary information, to second figure
Profile border as in is carried out after intensive treatment, the image after being handled, including：
According to the profile boundary information of all profiles in described first image, the profile of second image is extracted
The pixel on border, is obtained to the pixel on the profile border of second image according to equation below progress calculating
Take respective pixel point brightness：
Y2=K*Y1+b
Wherein：Y2 represents respective pixel point brightness, and Y1 represents the pixel after increasing quantification gradation processing
Brightness, K represents the Any Digit between 0-1, and b is any value between 0-255；
Increase luminance quantization grade is carried out to respective pixel point using equation below to handle, and is completed to described second
The intensive treatment of profile boundary luminances in image, it is ensured that luminance quantization of pixel of second image etc.
The uniformity of level：
Y3=[Y2/Q] * Q
Wherein：Y2 represents respective pixel point brightness, and Y3 is represented to the pixel after respective pixel point Y2 processing
Point brightness, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, [Y2/Q]
Represent and floor operation is done to Y2 and Q business.
The embodiment of the present invention also provides a kind of image processing apparatus, including：
Determining module is obtained, for obtaining pending image, the first area in the pending image is determined,
The first area is the region comprising default object or a predeterminable area；
First processing module, for carrying out brightness Fuzzy Processing to the pending image, obtains the first image；
Calculating processing module, the profile for extracting first area described in described first image, and according to
The profile of the first area, calculates the brightness judgement threshold on profile border, and differentiate according to the brightness
Thresholding, determines the profile boundary information of all profiles in described first image；
Second processing module, the thick quantification treatment for carrying out monochrome information to the pending image, is obtained
Second image；
Acquisition module is handled, for the profile boundary information using all profiles in described first image, to institute
The profile border stated in the second image is carried out after intensive treatment, the image after being handled.
Wherein, the acquisition determining module includes：
Judging submodule, for judging in the pending image with the presence or absence of default object；
First determination sub-module, for when there is default object in the pending image, it is determined that default pair
As shared region is the first area；
Second determination sub-module, for when default object is not present in the pending image, it is determined that default
Region is the first area.
Wherein, first determination sub-module includes：
First module, for when there is multiple default objects in the pending image, choosing area maximum
Default object, and determine that the region shared by the maximum default object of area is the first area；
Second unit, for when there are two size identicals in the pending image and preset object,
The default object near with centre coordinate distance is chosen, and is determined shared by the default object near with centre coordinate distance
Region be the first area.
Wherein, the calculating processing module includes：
Statistic submodule is extracted, the wire-frame image vegetarian refreshments for extracting the first area counts firstth area
Brightness maxima, brightness minimum value or the average brightness of the wire-frame image vegetarian refreshments in domain；
3rd determination sub-module, it is minimum for the product according to brightness maxima and the first proportionality coefficient, brightness
Value and the product or average brightness of the second proportionality coefficient determine the brightness judgement threshold on profile border, wherein
First proportionality coefficient is small 1 positive number, and second proportionality coefficient is the positive number more than 1.
Wherein, the Second processing module is further used for：
The pixel of the pending image is extracted, to the pixel of the pending image according to equation below
The grade processing of increase luminance quantization is carried out, second image is obtained：
Y1=[y/Q] * Q
Wherein：Y1 represents the pixel brightness after increasing quantification gradation processing, and y represents original pixel and lighted
Degree, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, and [y/Q] is represented to y
Floor operation is done with Q business.
Wherein, the processing acquisition module includes：
Extraction process submodule, for the profile boundary information according to all profiles in described first image, is carried
The pixel on the profile border of second image is taken, to the pixel root on the profile border of second image
According to equation below calculate and obtain respective pixel point brightness：
Y2=K*Y1+b
Wherein：Y2 represents respective pixel point brightness, and Y1 represents the pixel after increasing quantification gradation processing
Brightness, K represents the Any Digit between 0-1, and b is any value between 0-255；
Submodule is handled, is handled for carrying out increase luminance quantization grade to respective pixel point using equation below,
Complete the intensive treatment to the profile boundary luminances in second image, it is ensured that the pixel of second image
The uniformity of the luminance quantization grade of point：
Y3=[Y2/Q] * Q
Wherein：Y2 represents respective pixel point brightness, and Y3 is represented to the pixel after respective pixel point Y2 processing
Point brightness, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, [Y2/Q]
Represent and floor operation is done to Y2 and Q business.
The above-mentioned technical proposal of the embodiment of the present invention at least includes following beneficial effect：
The profile of the first area of first image of the embodiment of the present invention by acquisition Jing Guo brightness Fuzzy Processing
Pixel information, calculates according to the wire-frame image vegetarian refreshments information of the first area of the first image and obtains profile border
Brightness judgement threshold, and according to brightness judgement threshold, determine the profile border letter of all profiles in the first image
Breath, then according to the profile boundary information of all profiles in the first image, the thick quantization to carrying out monochrome information
The profile border of the second image after processing carries out intensive treatment, and the filter that can improve image is taken pictures effect,
The problem of filter shooting effect pit of effectively solution conventional images is excessive.
Brief description of the drawings
Fig. 1 is image processing method of embodiment of the present invention schematic diagram one；
Fig. 2 is image processing method of embodiment of the present invention schematic diagram two；
Fig. 3 a are the schematic diagram one of prewitt operators；
Fig. 3 b are the schematic diagram two of prewitt operators；
Fig. 4 is image processing method of embodiment of the present invention schematic diagram three；
Fig. 5 is image processing method of embodiment of the present invention schematic diagram four；
Fig. 6 is image processing method overall flow schematic diagram of the embodiment of the present invention；
Fig. 7 is image processing apparatus schematic diagram of the embodiment of the present invention.
Embodiment
To make the technical problem to be solved in the present invention, technical scheme and advantage clearer, below in conjunction with attached
Figure and specific embodiment are described in detail.
As shown in figure 1, the embodiment of the present invention provides a kind of image processing method, including：
S101, obtain pending image, determine the first area in pending image, first area be comprising
The region of default object or a predeterminable area；
S102, to pending image carry out brightness Fuzzy Processing, obtain the first image；
S103, the profile for extracting first area in the first image, and according to the profile of first area, calculate
Go out the brightness judgement threshold on profile border, and according to brightness judgement threshold, determine all profiles in the first image
Profile boundary information；
S104, the thick quantification treatment that monochrome information is carried out to pending image, obtain the second image；
S105, the profile boundary information using all profiles in the first image, to the profile side in the second image
Boundary is carried out after intensive treatment, the image after being handled.
Above-mentioned steps of the embodiment of the present invention, by using contours extract mechanism, increase luminance quantization level mechanism
The filter for improving image is taken pictures effect, effectively solve conventional images filter shooting effect pit it is excessive the problem of.
Specifically, pending image is obtained first, can be by the pending image of acquisition for the ease of calculating
Stored with yuv forms.It is then determined that the first area in pending image, first area here can
To be the region comprising default object or a predeterminable area.The default object of the present invention is with face figure
As exemplified by, those skilled in the art can also determine other default objects according to demand.
It should be noted that pending image is located in a coordinate system, when first area is a predeterminable area
When, it is generally the case that as shown in Fig. 2 predeterminable area is the region centered on the origin of coordinates, and in advance
If the size in region is pre-set, h represents the fore-and-aft distance of predeterminable area, and w represents predeterminable area
Lateral separation.
It is determined that, it is necessary to carry out the fuzzy place of brightness to pending image after first area in pending image
Reason, gets the first image, wherein when carrying out brightness Fuzzy Processing to pending image, can be using height
This filtering mode is handled.But the mode of Fuzzy Processing is not limited to gaussian filtering one kind, this area skill
Art personnel can be according to demand using other Fuzzy Processing modes.
, it is necessary to extract the profile of first area in the first image after the first image is acquired, and root
According to the profile of the first area extracted, using corresponding arithmetic operators, such as sobel or prewitt
Deng operator, the boundary information of first area is calculated, profile border is calculated according to the boundary information of first area
Brightness judgement threshold.Wherein the schematic diagram of prewitt operators is as best shown in figures 3 a and 3b.
According to obtained brightness judgement threshold, the profile boundary information of all profiles in the first image is determined.So
The thick quantification treatment of monochrome information is carried out for pending image afterwards, the second image is obtained.According to the first image
In all profiles profile boundary information, in the second image profile border carry out intensive treatment, obtain everywhere
Effect image after reason.It should be noted that the purpose for carrying out brightness intensive treatment to border is：Can be with
The profile of image is clearly distinguished out so that image effect is more preferable after processing.
The profile of the first area of first image of the embodiment of the present invention by acquisition Jing Guo brightness Fuzzy Processing
Pixel information, calculates according to the wire-frame image vegetarian refreshments information of the first area of the first image and obtains profile border
Brightness judgement threshold, and according to brightness judgement threshold, determine the profile border letter of all profiles in the first image
Breath, then according to the profile boundary information of all profiles in the first image, the thick quantization to carrying out monochrome information
The profile border of the second image after processing carries out intensive treatment, and the filter that can improve image is taken pictures effect,
The problem of filter shooting effect pit of effectively solution conventional images is excessive.
In the above embodiment of the present invention, the first area in pending image is determined in step S101, such as
Shown in Fig. 4, including：
S1011, judge in pending image with the presence or absence of default object；
S1012, when there is default object in pending image, it is determined that region shared by default object is the
One region；
S1013, when default object is not present in pending image, it is first area to determine predeterminable area.
Specifically, it is determined that during first area in pending image, it is necessary to judge in pending image whether
In the presence of default object.In the present embodiment, default object mentioned here is facial image.When pending
When there is facial image in image, i.e., when there is default object in pending image, determine shared by facial image
Region be first area.When facial image is not present in pending image, it is thus necessary to determine that predeterminable area is made
For first area.It should be noted that the position of predeterminable area and the size of predeterminable area are pre-
First set, those skilled in the art can be adjusted during pre-setting to it according to the actual requirements.
When there is default object in pending image, default object here is by taking facial image as an example, at this moment
Need to be divided into following several situations to determine first area：
When only existing a facial image in pending image, the area shared by facial image at this moment can be determined
Domain is first area；
, at this moment can be to the areas of multiple facial images when only existing multiple facial images in pending image
Size is compared, and determines the maximum facial image of area, then the area shared by the maximum facial image of area
Domain is first area；
When there is two size identical facial images in pending image, at this moment need to check face
The central point in region shared by image is with the origin of coordinates of the coordinate system where pending image apart from size.When
The distance between the central point in region shared by first facial image and the origin of coordinates are less than the second facial image institute
When the distance between the central point and the origin of coordinates in occupied area domain, it may be determined that region shared by the first facial image is
First area.
In the above embodiment of the present invention, the profile of first area in the first image is extracted in step S103,
And according to the profile of first area, the brightness judgement threshold on profile border is calculated, as shown in figure 5, including：
S1031, the wire-frame image vegetarian refreshments for extracting first area, count the brightness of the wire-frame image vegetarian refreshments of first area
Maximum, brightness minimum value or average brightness；
S1032, the product according to brightness maxima and the first proportionality coefficient, brightness minimum value and the second ratio
The product or average brightness of coefficient determine the brightness judgement threshold on profile border, wherein the first proportionality coefficient
For small 1 positive number, the second proportionality coefficient is the positive number more than 1.
Specifically, in the profile according to first area, the brightness judgement threshold on calculating profile border, first
The wire-frame image vegetarian refreshments of acquisition first area is needed, the wire-frame image vegetarian refreshments of the first area acquired is directed to,
The brightness value of each pixel is counted, determines that the brightness on profile border is sentenced according to the brightness value of each pixel
Other thresholding.
Determine the brightness judgement threshold on profile border in the brightness value according to each pixel, including but not office
It is limited to following several method：
The brightness value of each pixel is counted, and chooses brightness maxima, by the brightness maxima acquired
It is multiplied with the first proportionality coefficient, the brightness judgement threshold on profile border is obtained, wherein the first proportionality coefficient is
Any number between 0~1.
The brightness value of each pixel is counted, and chooses brightness minimum value, by the brightness minimum value acquired
It is multiplied with the second proportionality coefficient, the brightness judgement threshold on profile border is obtained, wherein the second proportionality coefficient is big
In 1 number.
The brightness value of each pixel is counted, and calculates the average value of the brightness value of each pixel, will be calculated
Obtained average value as profile border brightness judgement threshold.
The brightness value of each pixel is counted, and chooses brightness maxima.Then the bright of each pixel is calculated
The average value of angle value, takes the intermediate value of brightness maxima and average brightness, and the brightness for obtaining profile border differentiates
Thresholding.
After the brightness judgement threshold on profile border is acquired, it can be used according to brightness judgement threshold
Profile boundary extraction algorithm, determines the profile boundary information of all profiles in the first image.It should be noted that
Adopted when the operator used when using profile boundary extraction algorithm and the boundary information of calculating first area
Operator is identical.And brightness judgement threshold is herein as a restrictive condition so that in the first image
The pixel brightness on all profile borders is less than brightness judgement threshold.
In the above embodiment of the present invention, step S104 is further used for：
The pixel of pending image is extracted, the pixel of pending image is increased according to equation below
The processing of luminance quantization grade, obtains the second image：
Y1=[y/Q] * Q
Wherein：Y1 represents the pixel brightness after increasing quantification gradation processing, and y represents original pixel and lighted
Degree, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, and [y/Q] is represented to y
Floor operation is done with Q business.
Specifically, the process of the second image is obtained, it is necessary to obtain each pixel of pending image first,
The brightness for being directed to each pixel of pending image carries out the processing of increase luminance quantization grade.Specifically
Process is as follows：
Each pixel of pending image is directed to, increase luminance quantization etc. is carried out using equation below one
The processing of level.
Y1=[y/Q] * Q formula one
Wherein, Y1 represents the pixel brightness after increasing quantification gradation processing, and y represents original pixel and lighted
Degree, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, and [y/Q] is represented to y
Floor operation is done with Q business.
When being operated to original pixel point brightness y, it is thus necessary to determine that a quantification gradation, Q is quantification gradation
Numerical value, original pixel point brightness y span is 0~255, it is assumed that y value is that 25, Q value is
4, then after being calculated using formula one, the value for obtaining Y1 is 24.Specifically calculating process is：To 25
Divided by 4 business's progress floor operation obtains 6, obtained numerical value 24 that 6 are multiplied with 4 is to increase to quantify
Pixel brightness Y1 after grade processing.
Or assume that the value that y value is 23, Q is 4, then after being calculated using formula one, obtain
Y1 value is 20.Specifically calculating process is：5 are obtained to 23 divided by 4 business's progress floor operation,
Obtained numerical value 20 that 5 are multiplied with 4 is to increase the pixel brightness Y1 after quantification gradation processing.
After the pixel brightness Y1 after increasing quantification gradation processing is obtained, according to the picture after each renewal
Vegetarian refreshments obtains the second image.
In the above embodiment of the present invention, the profile side of all profiles in the first image is utilized in step S105
Boundary's information, is carried out after intensive treatment, the image after being handled to the profile border in the second image, including：
According to the profile boundary information of all profiles in the first image, the picture on the profile border of the second image is extracted
Vegetarian refreshments, carries out calculating acquisition respective pixel point to the pixel on the profile border of the second image according to equation below
Brightness：
Y2=K*Y1+b
Wherein：Y2 represents respective pixel point brightness, and Y1 represents the pixel after increasing quantification gradation processing
Brightness, K represents the Any Digit between 0-1, and b is any value between 0-255；
Increase luminance quantization grade is carried out to respective pixel point using equation below to handle, and is completed to the second image
In profile boundary luminances intensive treatment, it is ensured that the luminance quantization grade of the pixel of the second image it is consistent
Property：
Y3=[Y2/Q] * Q
Wherein：Y2 represents respective pixel point brightness, and Y3 is represented to the pixel after respective pixel point Y2 processing
Point brightness, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, [Y2/Q]
Represent and floor operation is done to Y2 and Q business.
Specifically, extracting the second image according to the profile boundary information of all profiles in the first image first
The pixel on profile border, brightness processed is carried out using the pixel on the profile border of two pair of second image of formula.
Y2=K*Y1+b formula two
It should be noted that when the pixel on the profile border of recycling two pair of second image of formula is handled,
Because Y2 span is 0~255, therefore, it is necessary to ensure to obtain in K or b sampling processes
Y2 numerical value not over 255.
0.1, b value is taken to take any number in 0~253 it is assumed that Y1 value is 20, K, in reality
In the calculating process of border, same image is directed to, K and b values are to determine.For example：K takes 0.1, b's
Value takes 4, Y1 value to be 20, is calculated according to formula two, obtained Y2 value is 6.It is directed to same
The different pixels point of one image, Y1 value is 30, is calculated according to formula two, obtaining Y2 value is
7.It should be noted that for an image, K and b values pre-set determination, according to the K of determination
With b values, calculated using formula two, obtained Y2 can be more than Y1, might be less that Y1, its
Middle Y2 and Y1 magnitude relationship are depended on, K and b value.When K takes 0.5, b value to take 25,
Y1 value is 30, is calculated according to formula two, the value for obtaining Y2 is 40, and now Y2 is more than Y1.
Recycle two pair of second image of formula profile border pixel handle after, it is necessary to utilize formula
Three couples of Y2 carry out the grade processing of increase luminance quantization.
Y3=[Y2/Q] * Q formula three
It should be noted that the quantification gradation used herein and using when handling pending image
Quantification gradation is identical, it is therefore intended that ensure the uniformity of the luminance quantization grade of the pixel of the second image.
For example：It is assumed that the value that Y2 value is 23, Q is 4, then after being calculated using formula three, obtain
Y3 value is 20.
As shown in fig. 6, being overall flow schematic diagram of the embodiment of the present invention：
S201, the pending image of acquisition；
S202, determine first area；
S203, brightness Fuzzy Processing is carried out to pending image obtain the first image；
S204, the profile for extracting first area in the first image, and according to the profile of first area, calculate
Go out the brightness judgement threshold on profile border；
S205, according to brightness judgement threshold, determine the profile boundary information of all profiles in the first image；
S206, the thick quantification treatment that monochrome information is carried out to pending image, obtain the second image；
S207, the profile boundary information using all profiles in the first image, to the profile side in the second image
Boundary carries out intensive treatment；
S208, output effect image.
The embodiment of the present invention also provides a kind of image processing apparatus, as shown in fig. 7, comprises：
Determining module 10 is obtained, for obtaining pending image, the first area in pending image is determined,
First area is the region comprising default object or a predeterminable area；
First processing module 20, for carrying out brightness Fuzzy Processing to pending image, obtains the first image；
Calculating processing module 30, the profile for extracting first area in the first image, and according to first
The profile in region, calculates the brightness judgement threshold on profile border, and according to brightness judgement threshold, determines
The profile boundary information of all profiles in one image；
Second processing module 40, the thick quantification treatment for carrying out monochrome information to pending image, is obtained
Second image；
Acquisition module 50 is handled, for the profile boundary information using all profiles in the first image, to the
Profile border in two images is carried out after intensive treatment, the image after being handled.
Wherein, obtaining determining module 10 includes：
Judging submodule 11, for judging in pending image with the presence or absence of default object；
First determination sub-module 12, for when there is default object in pending image, it is determined that default pair
As shared region is first area；
Second determination sub-module 13, for when default object is not present in pending image, it is determined that default
Region is first area.
Wherein, the first determination sub-module 12 includes：
First module 121, for when there is multiple default objects in pending image, choosing area maximum
Default object, and determine that the region shared by the maximum default object of area is first area；
Second unit 122, for when there are two size identicals in pending image and preset object,
The default object near with centre coordinate distance is chosen, and is determined shared by the default object near with centre coordinate distance
Region be first area.
Wherein, calculating processing module 30 includes：
Statistic submodule 31 is extracted, the wire-frame image vegetarian refreshments for extracting first area counts first area
Brightness maxima, brightness minimum value or the average brightness of wire-frame image vegetarian refreshments；
3rd determination sub-module 32, for the product according to brightness maxima and the first proportionality coefficient, brightness
Minimum value and the product or average brightness of the second proportionality coefficient determine the brightness judgement threshold on profile border,
Wherein the first proportionality coefficient is small 1 positive number, and the second proportionality coefficient is the positive number more than 1.
Wherein, Second processing module 40 is further used for：
The pixel of pending image is extracted, the pixel of pending image is increased according to equation below
The processing of luminance quantization grade, obtains the second image：
Y1=[y/Q] * Q
Wherein：Y1 represents the pixel brightness after increasing quantification gradation processing, and y represents original pixel and lighted
Degree, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, and [y/Q] is represented to y
Floor operation is done with Q business.
Wherein, processing acquisition module 50 includes：
Extraction process submodule 51, for the profile boundary information according to all profiles in the first image, is carried
The pixel on the profile border of the second image is taken, to the pixel on the profile border of the second image according to following public
Formula, which calculate, obtains respective pixel point brightness：
Y2=K*Y1+b
Wherein：Y2 represents respective pixel point brightness, and Y1 represents the pixel after increasing quantification gradation processing
Brightness, K represents the Any Digit between 0-1, and b is any value between 0-255；
Submodule 52 is handled, for carrying out increase luminance quantization grade to respective pixel point using equation below
Processing, completes the intensive treatment to the profile boundary luminances in the second image, it is ensured that the pixel of the second image
Luminance quantization grade uniformity：
Y3=[Y2/Q] * Q
Wherein：Y2 represents respective pixel point brightness, and Y3 is represented to the pixel after respective pixel point Y2 processing
Point brightness, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, [Y2/Q]
Represent and floor operation is done to Y2 and Q business.
The image processing method of the embodiment of the present invention, by obtaining the first image Jing Guo brightness Fuzzy Processing
The wire-frame image vegetarian refreshments information of first area, is calculated according to the wire-frame image vegetarian refreshments information of the first area of the first image
The brightness judgement threshold on profile border is obtained, and according to brightness judgement threshold, determines all wheels in the first image
Wide profile boundary information, it is bright to carrying out then according to the profile boundary information of all profiles in the first image
The profile border for the second image spent after the thick quantification treatment of information carries out intensive treatment, can improve image
Filter is taken pictures effect, effectively solve conventional images filter shooting effect pit it is excessive the problem of.
It should be noted that image processing apparatus provided in an embodiment of the present invention is the device using the above method,
Then all embodiments of the above method are applied to the device, and can reach same or analogous beneficial effect.
Described above is the preferred embodiment of the present invention, it is noted that for the common skill of the art
For art personnel, on the premise of principle of the present invention is not departed from, some improvements and modifications can also be made,
These improvements and modifications also should be regarded as protection scope of the present invention.
Claims (12)
1. a kind of image processing method, it is characterised in that including：
Pending image is obtained, the first area in the pending image is determined, the first area is bag
Region or a predeterminable area containing default object；
Brightness Fuzzy Processing is carried out to the pending image, the first image is obtained；
The profile of first area described in described first image is extracted, and according to the profile of the first area,
The brightness judgement threshold on profile border is calculated, and according to the brightness judgement threshold, determines first figure
The profile boundary information of all profiles as in；
The thick quantification treatment of monochrome information is carried out to the pending image, the second image is obtained；
Using the profile boundary information of all profiles in described first image, to the profile in second image
Border is carried out after intensive treatment, the image after being handled.
2. the method as described in claim 1, it is characterised in that in the determination pending image
First area, including：
Judge in the pending image with the presence or absence of default object；
When there is default object in the pending image, it is determined that region shared by default object is described the
One region；
When default object is not present in the pending image, it is the first area to determine predeterminable area.
3. method as claimed in claim 2, it is characterised in that described to exist when in the pending image
During default object, it is determined that the region shared by default object is the first area, including：
When there is multiple default objects in the pending image, the maximum default object of area is chosen, and
It is the first area to determine the region shared by the maximum default object of area；
When presetting object in the presence of two size identicals in the pending image, choose and sat with center
The default object of subject distance closely, and determine that the region shared by the default object near with centre coordinate distance is described
First area.
4. the method as described in claim 1, it is characterised in that described to extract institute in described first image
The profile of first area is stated, and according to the profile of the first area, the brightness for calculating profile border differentiates
Thresholding, including：
The wire-frame image vegetarian refreshments of the first area is extracted, the brightness of the wire-frame image vegetarian refreshments of the first area is counted
Maximum, brightness minimum value or average brightness；
Multiplied according to product, the brightness minimum value of brightness maxima and the first proportionality coefficient with the second proportionality coefficient
Product or average brightness determine the brightness judgement threshold on profile border, wherein first proportionality coefficient is small
1 positive number, second proportionality coefficient is the positive number more than 1.
5. the method as described in claim 1, it is characterised in that described bright to the pending image progress
The thick quantification treatment of information is spent, the second image is obtained, including：
The pixel of the pending image is extracted, to the pixel of the pending image according to equation below
The grade processing of increase luminance quantization is carried out, second image is obtained：
Y1=[y/Q] * Q
Wherein：Y1 represents the pixel brightness after increasing quantification gradation processing, and y represents original pixel and lighted
Degree, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, and [y/Q] is represented to y
Floor operation is done with Q business.
6. method as claimed in claim 5, it is characterised in that own in the utilization described first image
The profile boundary information of profile, carries out after intensive treatment to the profile border in second image, obtains everywhere
Image after reason, including：
According to the profile boundary information of all profiles in described first image, the profile of second image is extracted
The pixel on border, is obtained to the pixel on the profile border of second image according to equation below progress calculating
Take respective pixel point brightness：
Y2=K*Y1+b
Wherein：Y2 represents respective pixel point brightness, and Y1 represents the pixel after increasing quantification gradation processing
Brightness, K represents the Any Digit between 0-1, and b is any value between 0-255；
Increase luminance quantization grade is carried out to respective pixel point using equation below to handle, and is completed to described second
The intensive treatment of profile boundary luminances in image, it is ensured that luminance quantization of pixel of second image etc.
The uniformity of level：
Y3=[Y2/Q] * Q
Wherein：Y2 represents respective pixel point brightness, and Y3 is represented to the pixel after respective pixel point Y2 processing
Point brightness, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, [Y2/Q]
Represent and floor operation is done to Y2 and Q business.
7. a kind of image processing apparatus, it is characterised in that including：
Determining module is obtained, for obtaining pending image, the first area in the pending image is determined,
The first area is the region comprising default object or a predeterminable area；
First processing module, for carrying out brightness Fuzzy Processing to the pending image, obtains the first image；
Calculating processing module, the profile for extracting first area described in described first image, and according to
The profile of the first area, calculates the brightness judgement threshold on profile border, and differentiate according to the brightness
Thresholding, determines the profile boundary information of all profiles in described first image；
Second processing module, the thick quantification treatment for carrying out monochrome information to the pending image, is obtained
Second image；
Acquisition module is handled, for the profile boundary information using all profiles in described first image, to institute
The profile border stated in the second image is carried out after intensive treatment, the image after being handled.
8. device as claimed in claim 7, it is characterised in that the acquisition determining module includes：
Judging submodule, for judging in the pending image with the presence or absence of default object；
First determination sub-module, for when there is default object in the pending image, it is determined that default pair
As shared region is the first area；
Second determination sub-module, for when default object is not present in the pending image, it is determined that default
Region is the first area.
9. device as claimed in claim 8, it is characterised in that first determination sub-module includes：
First module, for when there is multiple default objects in the pending image, choosing area maximum
Default object, and determine that the region shared by the maximum default object of area is the first area；
Second unit, for when there are two size identicals in the pending image and preset object,
The default object near with centre coordinate distance is chosen, and is determined shared by the default object near with centre coordinate distance
Region be the first area.
10. device as claimed in claim 7, it is characterised in that the calculating processing module includes：
Statistic submodule is extracted, the wire-frame image vegetarian refreshments for extracting the first area counts firstth area
Brightness maxima, brightness minimum value or the average brightness of the wire-frame image vegetarian refreshments in domain；
3rd determination sub-module, it is minimum for the product according to brightness maxima and the first proportionality coefficient, brightness
Value and the product or average brightness of the second proportionality coefficient determine the brightness judgement threshold on profile border, wherein
First proportionality coefficient is small 1 positive number, and second proportionality coefficient is the positive number more than 1.
11. device as claimed in claim 7, it is characterised in that the Second processing module is further used
In：
The pixel of the pending image is extracted, to the pixel of the pending image according to equation below
The grade processing of increase luminance quantization is carried out, second image is obtained：
Y1=[y/Q] * Q
Wherein：Y1 represents the pixel brightness after increasing quantification gradation processing, and y represents original pixel and lighted
Degree, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, and [y/Q] is represented to y
Floor operation is done with Q business.
12. device as claimed in claim 11, it is characterised in that the processing acquisition module includes：
Extraction process submodule, for the profile boundary information according to all profiles in described first image, is carried
The pixel on the profile border of second image is taken, to the pixel root on the profile border of second image
According to equation below calculate and obtain respective pixel point brightness：
Y2=K*Y1+b
Wherein：Y2 represents respective pixel point brightness, and Y1 represents the pixel after increasing quantification gradation processing
Brightness, K represents the Any Digit between 0-1, and b is any value between 0-255；
Submodule is handled, is handled for carrying out increase luminance quantization grade to respective pixel point using equation below,
Complete the intensive treatment to the profile boundary luminances in second image, it is ensured that the pixel of second image
The uniformity of the luminance quantization grade of point：
Y3=[Y2/Q] * Q
Wherein：Y2 represents respective pixel point brightness, and Y3 is represented to the pixel after respective pixel point Y2 processing
Point brightness, Q represents the numerical value of quantification gradation, and Q value is any integer in 1-255, [Y2/Q]
Represent and floor operation is done to Y2 and Q business.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610067209.9A CN107025633B (en) | 2016-01-29 | 2016-01-29 | Image processing method and device |
| PCT/CN2016/090460 WO2017128646A1 (en) | 2016-01-29 | 2016-07-19 | Image processing method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610067209.9A CN107025633B (en) | 2016-01-29 | 2016-01-29 | Image processing method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107025633A true CN107025633A (en) | 2017-08-08 |
| CN107025633B CN107025633B (en) | 2020-11-27 |
Family
ID=59397386
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610067209.9A Active CN107025633B (en) | 2016-01-29 | 2016-01-29 | Image processing method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN107025633B (en) |
| WO (1) | WO2017128646A1 (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109145768A (en) * | 2018-07-31 | 2019-01-04 | 北京旷视科技有限公司 | Obtain the method and device of the human face data with face character |
| CN111483803A (en) * | 2020-04-17 | 2020-08-04 | 湖南视比特机器人有限公司 | Control method, capture system and storage medium |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113129218A (en) * | 2019-12-30 | 2021-07-16 | Tcl集团股份有限公司 | Method and terminal for processing image |
| CN114387500A (en) * | 2020-10-16 | 2022-04-22 | 苏州科瓴精密机械科技有限公司 | Image recognition method and system applied to self-walking device, self-walking device and readable storage medium |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101170641A (en) * | 2007-12-05 | 2008-04-30 | 北京航空航天大学 | A method for image edge detection based on threshold sectioning |
| CN101321289A (en) * | 2008-06-13 | 2008-12-10 | 北京大学 | Method, system and device for processing video image in mobile phone television |
| CN101727658A (en) * | 2008-10-14 | 2010-06-09 | 深圳迈瑞生物医疗电子股份有限公司 | Image processing method and device |
| CN101981582A (en) * | 2008-04-04 | 2011-02-23 | 富士胶片株式会社 | Method, apparatus, and program for detecting object |
| CN102612697A (en) * | 2009-07-21 | 2012-07-25 | 高通股份有限公司 | A method and system for detection and enhancement of video images |
| CN102920477A (en) * | 2012-03-05 | 2013-02-13 | 杭州弘恩医疗科技有限公司 | Device and method for determining target region boundary of medical image |
| US20130170756A1 (en) * | 2011-12-28 | 2013-07-04 | Dwango Co., Ltd. | Edge detection apparatus, program and method for edge detection |
| CN103514588A (en) * | 2013-10-10 | 2014-01-15 | 广东威创视讯科技股份有限公司 | Image enhancement method and system |
| US20140147057A1 (en) * | 2012-11-27 | 2014-05-29 | Texas Instruments Incorporated | Content adaptive edge and detail enhancement for image and video processing |
| US20140233846A1 (en) * | 2013-02-18 | 2014-08-21 | Armorlink Sh Corp. | Method for auto-depicting trends in object contours |
| CN104915934A (en) * | 2015-06-15 | 2015-09-16 | 电子科技大学 | Grayscale image enhancement method based on retina mechanism |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002247611A (en) * | 2001-02-16 | 2002-08-30 | Matsushita Electric Ind Co Ltd | Image discriminating circuit, image quality correcting device, image discrimination method, image quality correction method and program |
| CN101241526A (en) * | 2007-02-08 | 2008-08-13 | 佛山市顺德区顺达电脑厂有限公司 | Palm moving electronic device image identification authentication method |
-
2016
- 2016-01-29 CN CN201610067209.9A patent/CN107025633B/en active Active
- 2016-07-19 WO PCT/CN2016/090460 patent/WO2017128646A1/en active Application Filing
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101170641A (en) * | 2007-12-05 | 2008-04-30 | 北京航空航天大学 | A method for image edge detection based on threshold sectioning |
| CN101981582A (en) * | 2008-04-04 | 2011-02-23 | 富士胶片株式会社 | Method, apparatus, and program for detecting object |
| CN101321289A (en) * | 2008-06-13 | 2008-12-10 | 北京大学 | Method, system and device for processing video image in mobile phone television |
| CN101727658A (en) * | 2008-10-14 | 2010-06-09 | 深圳迈瑞生物医疗电子股份有限公司 | Image processing method and device |
| CN102612697A (en) * | 2009-07-21 | 2012-07-25 | 高通股份有限公司 | A method and system for detection and enhancement of video images |
| US20130170756A1 (en) * | 2011-12-28 | 2013-07-04 | Dwango Co., Ltd. | Edge detection apparatus, program and method for edge detection |
| CN102920477A (en) * | 2012-03-05 | 2013-02-13 | 杭州弘恩医疗科技有限公司 | Device and method for determining target region boundary of medical image |
| US20140147057A1 (en) * | 2012-11-27 | 2014-05-29 | Texas Instruments Incorporated | Content adaptive edge and detail enhancement for image and video processing |
| US20140233846A1 (en) * | 2013-02-18 | 2014-08-21 | Armorlink Sh Corp. | Method for auto-depicting trends in object contours |
| CN103514588A (en) * | 2013-10-10 | 2014-01-15 | 广东威创视讯科技股份有限公司 | Image enhancement method and system |
| CN104915934A (en) * | 2015-06-15 | 2015-09-16 | 电子科技大学 | Grayscale image enhancement method based on retina mechanism |
Non-Patent Citations (1)
| Title |
|---|
| 完颜丹丹: "红外图像对比度增强算法研究", 《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 |
|---|---|---|---|---|
| CN109145768A (en) * | 2018-07-31 | 2019-01-04 | 北京旷视科技有限公司 | Obtain the method and device of the human face data with face character |
| CN111483803A (en) * | 2020-04-17 | 2020-08-04 | 湖南视比特机器人有限公司 | Control method, capture system and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107025633B (en) | 2020-11-27 |
| WO2017128646A1 (en) | 2017-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107025633A (en) | A kind of image processing method and device | |
| Park et al. | Single image dehazing with image entropy and information fidelity | |
| Chen et al. | Depth image enhancement for Kinect using region growing and bilateral filter | |
| CN102231223B (en) | Foreign object blocking and alarming method used for camera | |
| US9811746B2 (en) | Method and system for detecting traffic lights | |
| CN104618700B (en) | Enhanced display method for color high dynamic range image | |
| CN107403413A (en) | A kind of video multiframe denoising and Enhancement Method | |
| CN107085707A (en) | A kind of license plate locating method based on Traffic Surveillance Video | |
| CN102722871A (en) | Quick and effective image enhancing method | |
| CN103679123B (en) | The method and system of analyte detection is left on a kind of ATM panels | |
| CN104952077B (en) | A kind of bill images greasy dirt detection method and system | |
| CN110569722A (en) | Visual analysis-based constructor dressing standard detection method and device | |
| CN109635758A (en) | Wisdom building site detection method is dressed based on the high altitude operation personnel safety band of video | |
| CN101739548A (en) | Eye detection method and system | |
| CN104574285A (en) | Method for automatically removing image black eyes | |
| CN105913385B (en) | Clarification method, system and the image processing apparatus of haze image | |
| CN104766096A (en) | Image classification method based on multi-scale global features and local features | |
| CN108921004A (en) | Safety cap wears recognition methods, electronic equipment, storage medium and system | |
| CN111275036A (en) | Target detection method, target detection device, electronic equipment and computer-readable storage medium | |
| CN104361335B (en) | A kind of processing method that black surround is automatically removed based on scan image | |
| CN107527374A (en) | A kind of method and apparatus of text importing | |
| CN110349090A (en) | A kind of image-scaling method based on newton second order interpolation | |
| CN102542304B (en) | Region segmentation skin-color algorithm for identifying WAP (Wireless Application Protocol) mobile porn image | |
| CN107274456A (en) | A kind of recognition methods of picture format and device | |
| CN109713745B (en) | A kind of intelligent charge detection circuit |
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 |