CN105069783A - Fuzzy picture identification method and device - Google Patents
Fuzzy picture identification method and device Download PDFInfo
- Publication number
- CN105069783A CN105069783A CN201510438070.XA CN201510438070A CN105069783A CN 105069783 A CN105069783 A CN 105069783A CN 201510438070 A CN201510438070 A CN 201510438070A CN 105069783 A CN105069783 A CN 105069783A
- Authority
- CN
- China
- Prior art keywords
- picture
- target photo
- pixel
- pixel value
- standard deviation
- 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 73
- 238000003708 edge detection Methods 0.000 claims description 19
- 230000000717 retained effect Effects 0.000 claims description 7
- 238000012805 post-processing Methods 0.000 claims description 6
- 238000012217 deletion Methods 0.000 claims description 5
- 230000037430 deletion Effects 0.000 claims description 5
- 238000012545 processing Methods 0.000 abstract description 4
- 238000009434 installation Methods 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 5
- 239000012141 concentrate Substances 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000002474 experimental method Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000006185 dispersion Substances 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
Classifications
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Landscapes
- Image Processing (AREA)
Abstract
The embodiment of the invention discloses a fuzzy picture identification method and a fuzzy picture identification device. A blurred picture recognition method can comprise the following steps: obtaining a target picture; obtaining a gray level picture corresponding to the target picture; carrying out sharpening processing on the gray level picture to obtain a sharpened picture; determining the pixel value of each pixel point of the sharpened picture; determining a standard deviation of the pixel value of the target picture according to the pixel value of each pixel point of the sharpened picture; and if the standard deviation is lower than a preset standard deviation threshold value, determining the target picture as a fuzzy picture. By applying the technical scheme provided by the embodiment of the invention, the process of identifying the fuzzy picture is simpler, and the identification rate of identifying a large number of pictures can be ensured.
Description
Technical field
The present invention relates to technical field of image processing, particularly a kind of blurred picture recognition methods and device.
Background technology
Nowadays, in the work and life of people, operable camera installation gets more and more, and as camera, mobile phone, panel computer etc. with camera, people can indiscriminately ad. as one wishes take pictures.Along with the accumulation of time, the picture obtained of taking pictures gets more and more, and the storage space of needs is increasing.In fact, in the process of taking pictures, fuzzy or unsharp picture can be produced unavoidably, if these pictures are all stored, more storage space will be wasted, and the storage space of the terminal such as camera installation or computer is all limited, in a lot of situation, need to delete these pictures or other process.
Delete these blurred pictures at present or before other process, need manually to identify which picture is fuzzy photo, if the quantity of picture is a lot, this process identified is comparatively loaded down with trivial details, and recognition rate is lower.
Summary of the invention
For solving the problem, the embodiment of the invention discloses a kind of blurred picture recognition methods and device.Technical scheme is as follows:
A kind of blurred picture recognition methods, comprising:
Obtain Target Photo;
Obtain the gray scale picture that described Target Photo is corresponding;
Edge contrast is carried out to described gray scale picture, obtains sharpening picture;
Determine the pixel value of each pixel of described sharpening picture;
According to the pixel value of each pixel of described sharpening picture, determine the standard deviation of the pixel value for described Target Photo;
If described Target Photo lower than the standard deviation threshold method preset, is then defined as blurred picture by described standard deviation.
In a kind of embodiment of the present invention, the pixel value of described each pixel according to described sharpening picture, determine the standard deviation of the pixel value for described Target Photo, comprising:
For each pixel of described sharpening picture, judge that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel;
Calculate the standard deviation of the pixel value of the pixel retained;
The standard deviation calculated is defined as the standard deviation of the pixel value for described Target Photo.
In a kind of embodiment of the present invention, the gray scale picture that the described Target Photo of described acquisition is corresponding, comprising:
According to the first proportionate relationship preset, process is reduced to described Target Photo, carrying out gray proces to reducing the picture after process, obtaining the gray scale picture that described Target Photo is corresponding;
Or,
Obtain and described Target Photo gray scale picture of the same size, according to the second proportionate relationship preset, process is reduced to obtained gray scale picture, obtains the gray scale picture that described Target Photo is corresponding.
In a kind of embodiment of the present invention, described described Target Photo is defined as blurred picture after, also comprise:
Export the information of whether deleting described Target Photo, according to the selection of user for described information, determine whether to perform the operation of deleting described Target Photo;
Or,
The described Target Photo of direct deletion;
Or,
Described Target Photo is put into default picture file folder to be deleted.
In a kind of embodiment of the present invention, described Edge contrast is carried out to described gray scale picture, obtains sharpening picture, comprising:
Use edge detection operator to carry out Edge contrast to described gray scale picture, obtain sharpening picture;
Wherein, described edge detection operator is the one in Laplce Laplacian operator, Sobel Sobel operator, Robert Roberts operator, triumphant Buddhist nun Canny operator.
A kind of blurred picture recognition device, comprising:
Target Photo obtains module, for obtaining Target Photo;
Gray scale picture obtains module, for obtaining gray scale picture corresponding to described Target Photo;
Sharpening picture obtains module, for carrying out Edge contrast to described gray scale picture, obtains sharpening picture;
Pixel value determination module, for determining the pixel value of each pixel of described sharpening picture;
Standard deviation determination module, for the pixel value of each pixel according to described sharpening picture, determines the standard deviation of the pixel value for described Target Photo, if described standard deviation is lower than the standard deviation threshold method preset, then triggers blurred picture determination module;
Described blurred picture determination module, for being defined as blurred picture by described Target Photo.
In a kind of embodiment of the present invention, described standard deviation determination module, comprising:
Judge submodule, for each pixel for described sharpening picture, judge that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel;
Calculating sub module, for calculating the standard deviation of the pixel value of the pixel of reservation;
Standard deviation determination submodule, for being defined as the standard deviation of the pixel value for described Target Photo by the standard deviation calculated.
In a kind of embodiment of the present invention, described gray scale picture obtains module, specifically for:
According to the first proportionate relationship preset, process is reduced to described Target Photo, carrying out gray proces to reducing the picture after process, obtaining the gray scale picture that described Target Photo is corresponding;
Or,
Obtain and described Target Photo gray scale picture of the same size, according to the second proportionate relationship preset, process is reduced to obtained gray scale picture, obtains the gray scale picture that described Target Photo is corresponding.
In a kind of embodiment of the present invention, also comprise post-processing module:
Described post-processing module, for after described Target Photo is defined as blurred picture, exports the information of whether deleting described Target Photo, according to the selection of user for described information, determines whether to perform the operation of deleting described Target Photo;
Or,
The described Target Photo of direct deletion;
Or,
Described Target Photo is put into default picture file folder to be deleted.
In a kind of embodiment of the present invention, described sharpening picture obtains module, specifically for:
Use edge detection operator to carry out Edge contrast to described gray scale picture, obtain sharpening picture;
Wherein, described edge detection operator is the one in Laplce Laplacian operator, Sobel Sobel operator, Robert Roberts operator, triumphant Buddhist nun Canny operator.
The technical scheme that the application embodiment of the present invention provides, after the gray scale picture corresponding to the Target Photo obtained carries out Edge contrast, according to the pixel value of each pixel of sharpening picture, determine the standard deviation of the pixel value of Target Photo, standard deviation is lower than the standard deviation threshold method preset, then can show that the object edge of this Target Photo is clear not, this Target Photo can be defined as blurred picture, identify that the process of blurred picture is comparatively simple, the recognition rate that a large amount of pictures is identified can be ensured.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, be briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is a kind of implementing procedure figure of blurred picture recognition methods in the embodiment of the present invention;
Fig. 2 is a kind of schematic diagram of gray scale picture in the embodiment of the present invention;
Fig. 3 is a kind of structural representation of blurred picture recognition device in the embodiment of the present invention.
Embodiment
Technical scheme in the embodiment of the present invention is understood better in order to make those skilled in the art, below in conjunction with the accompanying drawing in the embodiment of the present invention, technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making the every other embodiment obtained under creative work prerequisite, belong to the scope of protection of the invention.
Shown in Figure 1, be the implementing procedure figure of a kind of blurred picture recognition methods that the embodiment of the present invention provides, the method can comprise the following steps:
S110: obtain Target Photo;
Be understandable that, whether Target Photo is to be identified its is the picture of blurred picture.
The technical scheme that the embodiment of the present invention provides can be applied to client, can also be applied to server.
In actual applications, Target Photo can be client according to the determined picture of the recognition instruction of user, the identification information of Target Photo or the store path information of Target Photo can be comprised in the recognition instruction of user.Such as, when user arranges photo on a storage device, if send recognition instruction for the photo be stored in certain file to client, then the photo in this file can be defined as Target Photo by client one by one.
Or use camera installation to take pictures user, when obtaining a new photo, this photo is directly defined as Target Photo by client.
Or, user by client by the photo upload in memory device to server, the photo that user uploads is defined as Target Photo by server one by one.
Certainly, can also obtain Target Photo by other means, the embodiment of the present invention does not limit this.
S120: obtain the gray scale picture that described Target Photo is corresponding;
Gray proces is done to arbitrary pictures, the gray scale picture of this picture can be obtained.It should be noted that, those skilled in the art can according to the common practise of this area and common technology side's section, and carry out gray proces to picture, the present invention does not limit this.For example, for the picture of certain yuv format, directly according to the Y-component information of this picture, the gray scale picture of this picture can be obtained.
In a kind of embodiment of the present invention, according to the first proportionate relationship preset, process can be reduced to described Target Photo, carrying out gray proces to reducing the picture after process, obtaining the gray scale picture that described Target Photo is corresponding.
Such as, Target Photo is the picture of 512*512 size, first Target Photo can be reduced into the picture of 64*64 size, and then carries out gray proces to reducing the picture after process, can obtain the gray scale picture that this Target Photo is corresponding.
In another kind of embodiment of the present invention, first can obtain the gray scale picture of the same size with described Target Photo, again according to the second proportionate relationship preset, process is reduced to obtained gray scale picture, obtains the gray scale picture that described Target Photo is corresponding.
Such as, Target Photo is the picture of 512*512 size, after gray proces is carried out to this picture, the gray scale picture of the same size with this Target Photo can be obtained, obtained gray scale picture is reduced into the picture of 64*64 size, can using the gray scale picture after reducing as gray scale picture corresponding to this Target Photo.
Consider no matter the technical scheme that the embodiment of the present invention provides is in the client executing of end side or performs at server, all may be subject to the impact of the factor such as computing power and hardware performance, especially when the client executing of end side, impact is more obvious, so to Target Photo or and after Target Photo gray scale picture of the same size reduces process, carry out further operation again and can reduce calculated amount, improve the recognition rate of picture.
In theory, the pixel of Target Photo is more, and recognition accuracy is higher.In actual applications, for ensureing certain recognition accuracy, can from Target Photo or with Target Photo gray scale picture of the same size proportionally extract 64*x pixel, the length of the gray scale picture that the Target Photo namely obtained is corresponding or wide in the quantity of smaller's pixel be 64, in addition according to the quantity of its pixel of length breadth ratio be more than or equal to 64 a value.
It should be noted that, above-mentioned the first default proportionate relationship and the second proportionate relationship preset can carry out arranging and adjusting according to actual needs.
S130: carry out Edge contrast to described gray scale picture, obtains sharpening picture;
S140: the pixel value determining each pixel of described sharpening picture;
For convenience of description, above-mentioned two steps are combined be described.
Be understandable that, blurred picture has such common ground: object edge is unintelligible, and the quantity of the pixel of marginal portion intermediate color is many.After obtaining gray scale picture corresponding to Target Photo, although the color information of losing, feature and the fuzzy message of edge transition are still retained.Utilize the edge detection operator in graph image effectively can filter the marginal information of object in picture, the marginal information of object in gray scale picture can be extracted by edge detection operator.
In a kind of embodiment of the present invention, edge detection operator can be used to carry out Edge contrast to described gray scale picture, obtain sharpening picture; Wherein, described edge detection operator is the one in Laplce Laplacian operator, Sobel Sobel operator, Robert Roberts operator, triumphant Buddhist nun Canny operator.Certainly, operable edge detection operator is not limited to above-mentioned several, and the embodiment of the present invention does not limit this.
To use Laplace operator to carry out Edge contrast to described gray scale picture, obtaining sharpening picture is that example is described.
When using Laplace operator to carry out rim detection, there is direction to have nothing to do and dual edge characteristic, namely after Laplace operator process, in picture, the pixel at the edge of object presents corresponding positive and negative two class values, and the gray scale difference value of the pixel at the absolute value less explanation edge of this two class value is less.
For the picture of gray scale shown in Fig. 2, proportionally extract 16*x pixel in this gray scale picture, the pixel matrix of each pixel obtained, namely gray-scale map information is as follows:
[58,89,63,41,54,54,158,177,190,200,214,211,218,223,230,232]
[69,78,61,61,61,45,55,124,188,198,201,210,218,227,230,231]
[78,98,41,76,58,63,42,40,211,183,219,199,153,224,235,237]
[84,86,70,61,71,182,149,145,122,146,103,120,235,237,232,133]
[77,207,203,197,185,126,34,133,149,98,146,163,148,234,237,144]
[41,73,51,172,173,84,140,147,231,142,23,58,143,145,144,231]
[247,211,162,167,163,145,36,30,140,142,143,93,117,144,144,145]
[249,249,42,35,147,151,147,13,146,140,136,136,29,47,27,73]
[245,246,199,117,143,144,140,134,18,33,22,24,65,61,100,113]
[247,245,158,33,138,142,24,18,18,56,62,60,58,104,110,116]
[246,245,229,199,16,57,55,55,18,61,66,88,9,18,31,241]
[246,244,244,91,14,54,54,55,20,72,49,16,17,104,121,241]
[246,243,241,183,23,51,53,59,13,59,69,76,36,118,91,239]
[244,244,243,211,66,17,23,58,19,67,82,81,24,102,220,231]
[245,245,247,233,109,59,85,140,23,42,40,77,85,93,212,222]
[249,249,248,238,205,61,60,13,26,71,73,102,155,172,177,212]
[250,251,249,245,226,22,61,63,32,113,148,150,163,170,173,186]
[252,252,251,248,233,55,59,104,137,145,150,159,163,165,166,177]
[251,250,248,248,237,221,126,131,139,147,154,159,160,166,168,173]
[246,248,247,246,239,227,178,135,146,154,160,158,165,167,169,171]
[239,241,245,243,238,226,203,168,152,153,155,162,163,166,167,168]
To the pixel value of each pixel in above-mentioned pixel matrix through Laplace operator pattern matrix
The numerical matrix obtained after computing is as follows:
[5,-1,-5,-26,53,149,-36,-29,-20,37,-11,-64,-13,3]
[-109,141,-83,39,75,139,362,-311,42,-190,-94,264,-44,-17]
[115,111,170,202,-319,-193,-136,163,-78,219,220,-282,-23,-86]
[-389,-287,-167,-173,-19,412,-57,-12,191,-197,-180,183,-169,-194]
[218,406,-100,-88,248,-259,-54,-346,-74,397,190,-104,85,181]
[-113,-177,-136,-20,-146,318,216,-11,-3,-178,82,-59,-123,-116]
[-248,477,333,-96,-21,-248,405,-273,-103,-103,-262,249,73,256]
[-46,-233,-58,-26,0,-111,-347,259,104,167,187,-88,72,-89]
[-84,74,480,-218,-205,259,159,38,-50,-44,-8,6,-169,-89]
[-16,-70,-427,344,39,-30,-74,82,-32,-4,-201,145,176,366]
[2,-171,276,128,-40,1,-32,78,-99,27,166,97,-142,-17]
[3,-51,-166,222,-57,-25,-57,105,-15,-10,-102,91,-139,334]
[-1,-29,-119,96,131,121,9,85,-66,-71,-65,208,47,-244]
[5,-19,-127,127,36,-58,-381,135,33,114,0,9,199,-136]
[-3,-9,-21,-186,102,-20,237,35,-30,69,47,-98,-93,61]
[-4,-1,-19,-199,315,-40,-42,211,-56,-106,-28,-14,-7,7]
[-4,-7,-15,-166,315,110,-26,-128,-33,6,-14,-5,5,19]
[-1,4,-13,-7,-239,85,-20,5,4,0,-5,13,-4,2]
[-8,-1,-7,-8,-44,-21,83,-4,-10,-19,14,-12,-2,-3]
Can find out that from upper group of pixel matrix the gray-scale value difference of the object edge part of picture is larger.Above-mentioned is for 16*x the explanation that pixel carries out, in actual applications, as previously mentioned, the computing power of terminal or server, hardware performance and recognition accuracy is considered, 64*x pixel can be chosen as statistical sample, to retain effective fuzzy message.After obtaining sharpening picture, the pixel value of each pixel of sharpening picture can be determined.
S150: according to the pixel value of each pixel of described sharpening picture, determine the standard deviation of the pixel value for described Target Photo;
In statistics, variance or standard deviation representative distribution dispersion degree, be worth less representative and distribute more concentrated, fluctuate less.In step S130, Edge contrast is carried out to gray scale picture, after obtaining the pixel value of each pixel of sharpening picture, variance or the standard deviation of the pixel value of pixel in this sharpening picture can be calculated, the pixel value difference be worth in less explanation gray-scale map sheet between pixel is less, the edge representing object in this Target Photo is clear not, and the standard deviation namely calculated can use the index that whether fuzzy judge Target Photo is.
In a kind of embodiment of the present invention, step S150 can comprise the following steps:
First step: for each pixel of described sharpening picture, judges that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel;
Second step: the standard deviation calculating the pixel value of the pixel retained;
3rd step: the standard deviation standard deviation calculated being defined as the pixel value for described Target Photo.
For convenience of description, above-mentioned three steps are combined be described.
In step S140, determine the pixel value of each pixel of sharpening picture.In actual applications, after edge detection operator computing, in sharpening picture, the pixel value of non-edge pixels point concentrates in certain numerical range, different edge detection operators is used to carry out computing to the gray scale picture of Target Photo, the pixel value of the non-edge pixels point obtained the numerical range concentrated may be different, can rule of thumb or a large amount of test result carry out checking and obtain this numerical range, after Laplace operator computing, the pixel value of non-edge pixels point concentrates on [-10, 10] between, the pixel value of these non-edge pixels points can be filtered out, namely these pixels can not be added up.For each pixel of sharpening picture, judge that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel.
To the pixel retained, calculate the standard deviation of the pixel value of these pixels, one of them that can use formula one or formula two calculates.
Formula one:
Wherein, σ is the pixel value x carrying out the pixel added up
1, x
2..., x
nstandard deviation, n is the pixel number of carrying out adding up, and x is the mean value of the pixel value carrying out the pixel added up.
Formula two:
σ＝sqrt(E(x
2)-[E(x)]
2)；
Wherein, σ is the standard deviation of the pixel value carrying out the pixel added up, E (x) for carrying out the expectation value of the pixel value of the pixel added up, E (x
2) for carry out the pixel value of the pixel added up square expectation value.
The standard deviation calculated can be defined as the standard deviation of the pixel value for Target Photo.
S160: if described standard deviation is lower than the standard deviation threshold method preset, then described Target Photo is defined as blurred picture.
Standard deviation threshold method can be obtained by a large amount of test experiments checking.
Such as, at same position shooting jobbie, after gray proces, Edge contrast are carried out to the clear pictures of this object, calculating pixel number is the standard deviation of the pixel value of the pixel of 5146 is 69.48, after carrying out gray proces, Edge contrast to the fuzzy photo of this object, calculating pixel number is the standard deviation of the pixel value of the pixel of 5146 is 25.53.
Or, gray proces, Edge contrast after filtering out non-edge pixels point are carried out to the clear pictures of this object, calculating pixel number is the standard deviation of the pixel value of the pixel of 2743 is 95.07, carry out gray proces, Edge contrast after filtering out edge pixel point to the fuzzy photo of this object, to calculate pixel number be the standard deviation of the pixel value of the pixel of 1971 is 40.79.
Use said method to carry out a large amount of test experiments, the difference limen that can settle the standard value is set as that 50 ~ 60 is proper, if namely this Target Photo, lower than the standard deviation threshold method of setting, can be defined as blurred picture by the standard deviation of the pixel value that Target Photo is corresponding.
Certainly, use different gray scale processing methods, different Edge contrast methods, the standard deviation threshold method obtained may be different, and this standard deviation threshold method can carry out setting and adjusting according to actual conditions.Fuzzy itself is also relative concept, and for the camera installation that image quality itself is not high, can suitably debase the standard difference limen value, or can arrange ladder threshold value according to the resolution of camera installation and image quality.
After Target Photo is defined as blurred picture, can operate further.
Such as, can mark this Target Photo, check for user.
Or, the information of whether deleting described Target Photo can be exported, according to the selection of user for described information, determine whether to perform the operation of deleting described Target Photo.If receive the delete instruction of user, then this Target Photo can be carried out delete processing.
Or, can directly delete the blurred picture determined.Once Target Photo is defined as blurred picture, directly carry out delete processing.
Or described Target Photo can be put into default picture file folder to be deleted, user can check the picture being defined as blurred picture in this file.
The technical scheme that the application embodiment of the present invention provides, after the gray scale picture corresponding to the Target Photo obtained carries out Edge contrast, according to the pixel value of each pixel of sharpening picture, determine the standard deviation of the pixel value of Target Photo, standard deviation is lower than the standard deviation threshold method preset, then can show that the object edge of this Target Photo is clear not, this Target Photo can be defined as blurred picture, identify that the process of blurred picture is comparatively simple, the recognition rate that a large amount of pictures is identified can be ensured.
Corresponding to said method embodiment, the embodiment of the present invention additionally provides a kind of blurred picture recognition device, shown in Figure 3, and this device can comprise with lower module:
Target Photo obtains module 310, for obtaining Target Photo;
Gray scale picture obtains module 320, for obtaining gray scale picture corresponding to described Target Photo;
Sharpening picture obtains module 330, for carrying out Edge contrast to described gray scale picture, obtains sharpening picture;
Pixel value determination module 340, for determining the pixel value of each pixel of described sharpening picture;
Standard deviation determination module 350, for the pixel value of each pixel according to described sharpening picture, determines the standard deviation of the pixel value for described Target Photo, if described standard deviation is lower than the standard deviation threshold method preset, then triggers blurred picture determination module 360;
Described blurred picture determination module 360, for being defined as blurred picture by described Target Photo.
Be understandable that, whether Target Photo is to be identified its is the picture of blurred picture.
The device that the embodiment of the present invention provides can be applied to client, can also be applied to server.
In actual applications, Target Photo can be client according to the determined picture of the recognition instruction of user, the identification information of Target Photo or the store path information of Target Photo can be comprised in the recognition instruction of user.Such as, when user arranges photo on a storage device, if send recognition instruction for the photo be stored in certain file to client, then the photo in this file can be defined as Target Photo by client one by one.
Or use camera installation to take pictures user, when obtaining a new photo, this photo is directly defined as Target Photo by client.
Or, user by client by the photo upload in memory device to server, the photo that user uploads is defined as Target Photo by server one by one.
Certainly, Target Photo obtains module can also obtain Target Photo by other means, and the embodiment of the present invention does not limit this.
Gray proces is done to arbitrary pictures, the gray scale picture of this picture can be obtained.It should be noted that, those skilled in the art can according to the common practise of this area and common technology side's section, and carry out gray proces to picture, the present invention does not limit this.For example, for the picture of certain yuv format, directly according to the Y-component information of this picture, the gray scale picture of this picture can be obtained.
Be understandable that, blurred picture has such common ground: object edge is unintelligible, and the quantity of the pixel of marginal portion intermediate color is many.After obtaining gray scale picture corresponding to Target Photo, although the color information of losing, feature and the fuzzy message of edge transition are still retained.Utilize the edge detection operator in graph image effectively can filter the marginal information of object in picture, the marginal information of object in gray scale picture can be extracted by edge detection operator.
In statistics, variance or standard deviation represent distribution consistency degree, are worth less representative and distribute more concentrated.Edge contrast is carried out to gray scale picture, after obtaining the pixel value of each pixel of sharpening picture, can calculate variance or the standard deviation of the pixel value of pixel in this sharpening picture, the edge being worth object in gray scale picture corresponding to this Target Photo of less explanation is clear not.If this Target Photo lower than the standard deviation threshold method preset, is then defined as blurred picture by standard deviation.
In a kind of embodiment of the present invention, described standard deviation determination module 350, can comprise following submodule:
Judge submodule, for each pixel for described sharpening picture, judge that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel;
Calculating sub module, for calculating the standard deviation of the pixel value of the pixel of reservation;
Standard deviation determination submodule, for being defined as the standard deviation of the pixel value for described Target Photo by the standard deviation calculated.
In actual applications, after edge detection operator computing, in sharpening picture, the pixel value of non-edge pixels point concentrates in certain numerical range, different edge detection operators is used to carry out computing to the gray scale picture of Target Photo, the pixel value of the non-edge pixels point obtained the numerical range concentrated may be different, can rule of thumb or a large amount of test result carry out checking and obtain this numerical range, after Laplace operator computing, the pixel value of non-edge pixels point concentrates on [-10, 10] between, the pixel value of these non-edge pixels points can be filtered out, namely these pixels can not be added up.For each pixel of sharpening picture, judge that submodule judges that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel.
To the pixel retained, calculating sub module calculates the standard deviation of the pixel value of these pixels.The standard deviation calculated can be defined as the standard deviation of the pixel value for Target Photo by standard deviation determination submodule.
In a kind of embodiment of the present invention, described gray scale picture obtains module 320, can be specifically for:
According to the first proportionate relationship preset, process is reduced to described Target Photo, carrying out gray proces to reducing the picture after process, obtaining the gray scale picture that described Target Photo is corresponding;
Or,
Obtain and described Target Photo gray scale picture of the same size, according to the second proportionate relationship preset, process is reduced to obtained gray scale picture, obtains the gray scale picture that described Target Photo is corresponding.
Consider no matter the device that the embodiment of the present invention provides is run in the client of end side or run at server, all may be subject to the impact of the factor such as computing power and hardware performance, especially when the client of end side is run, impact is more obvious, so to Target Photo or and after Target Photo gray scale picture of the same size reduces process, carry out further operation again and can reduce calculated amount, improve the recognition rate of picture.
In theory, the pixel of Target Photo is more, and recognition accuracy is higher.In actual applications, for ensureing certain recognition accuracy, can from Target Photo or with Target Photo gray scale picture of the same size proportionally extract 64*x pixel, the length of the gray scale picture that the Target Photo namely obtained is corresponding or wide in the quantity of smaller's pixel be 64, in addition according to the quantity of its pixel of length breadth ratio be more than or equal to 64 a value.
It should be noted that, above-mentioned the first default proportionate relationship and the second proportionate relationship preset can carry out arranging and adjusting according to actual needs.
In one embodiment of the invention, this device can also comprise post-processing module:
Described post-processing module, after described Target Photo is defined as blurred picture, exports the information of whether deleting described Target Photo, according to the selection of user for described information, determines whether to perform the operation of deleting described Target Photo;
Or,
The described Target Photo of direct deletion;
Or,
Described Target Photo is put into default picture file folder to be deleted.
In a kind of embodiment of the present invention, described sharpening picture obtains module 330, can be specifically for:
Use edge detection operator to carry out Edge contrast to described gray scale picture, obtain sharpening picture;
Wherein, described edge detection operator is the one in Laplce Laplacian operator, Sobel Sobel operator, Robert Roberts operator, triumphant Buddhist nun Canny operator.
The device that the application embodiment of the present invention provides, after the gray scale picture corresponding to the Target Photo obtained carries out Edge contrast, according to the pixel value of each pixel of sharpening picture, determine the standard deviation of the pixel value of Target Photo, standard deviation is lower than the standard deviation threshold method preset, then can show that the object edge of this Target Photo is clear not, this Target Photo can be defined as blurred picture, identify that the process of blurred picture is comparatively simple, the recognition rate that a large amount of pictures is identified can be ensured.
It should be noted that, in this article, the such as relational terms of first and second grades and so on is only used for an entity or operation to separate with another entity or operational zone, and not necessarily requires or imply the relation that there is any this reality between these entities or operation or sequentially.And, term " comprises ", " comprising " or its any other variant are intended to contain comprising of nonexcludability, thus make to comprise the process of a series of key element, method, article or equipment and not only comprise those key elements, but also comprise other key elements clearly do not listed, or also comprise by the intrinsic key element of this process, method, article or equipment.When not more restrictions, the key element limited by statement " comprising ... ", and be not precluded within process, method, article or the equipment comprising described key element and also there is other identical element.
Each embodiment in this instructions all adopts relevant mode to describe, between each embodiment identical similar part mutually see, what each embodiment stressed is the difference with other embodiments.Especially, for device embodiment, because it is substantially similar to embodiment of the method, so description is fairly simple, relevant part illustrates see the part of embodiment of the method.
One of ordinary skill in the art will appreciate that all or part of step realized in said method embodiment is that the hardware that can carry out instruction relevant by program has come, described program can be stored in computer read/write memory medium, here the alleged storage medium obtained, as: ROM/RAM, magnetic disc, CD etc.
The foregoing is only preferred embodiment of the present invention, be not intended to limit protection scope of the present invention.All any amendments done within the spirit and principles in the present invention, equivalent replacement, improvement etc., be all included in protection scope of the present invention.
Claims (10)
1. a blurred picture recognition methods, is characterized in that, comprising:
Obtain Target Photo;
Obtain the gray scale picture that described Target Photo is corresponding;
Edge contrast is carried out to described gray scale picture, obtains sharpening picture;
Determine the pixel value of each pixel of described sharpening picture;
According to the pixel value of each pixel of described sharpening picture, determine the standard deviation of the pixel value for described Target Photo;
If described Target Photo lower than the standard deviation threshold method preset, is then defined as blurred picture by described standard deviation.
2. method according to claim 1, is characterized in that, the pixel value of described each pixel according to described sharpening picture, determines the standard deviation of the pixel value for described Target Photo, comprising:
For each pixel of described sharpening picture, judge that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel;
Calculate the standard deviation of the pixel value of the pixel retained;
The standard deviation calculated is defined as the standard deviation of the pixel value for described Target Photo.
3. method according to claim 1, is characterized in that, the gray scale picture that the described Target Photo of described acquisition is corresponding, comprising:
According to the first proportionate relationship preset, process is reduced to described Target Photo, carrying out gray proces to reducing the picture after process, obtaining the gray scale picture that described Target Photo is corresponding;
Or,
Obtain and described Target Photo gray scale picture of the same size, according to the second proportionate relationship preset, process is reduced to obtained gray scale picture, obtains the gray scale picture that described Target Photo is corresponding.
4. method according to claim 1, is characterized in that, described described Target Photo is defined as blurred picture after, also comprise:
Export the information of whether deleting described Target Photo, according to the selection of user for described information, determine whether to perform the operation of deleting described Target Photo;
Or,
The described Target Photo of direct deletion;
Or,
Described Target Photo is put into default picture file folder to be deleted.
5. the method according to any one of Claims 1-4, is characterized in that, describedly carries out Edge contrast to described gray scale picture, obtains sharpening picture, comprising:
Use edge detection operator to carry out Edge contrast to described gray scale picture, obtain sharpening picture.
6. a blurred picture recognition device, is characterized in that, comprising:
Target Photo obtains module, for obtaining Target Photo;
Gray scale picture obtains module, for obtaining gray scale picture corresponding to described Target Photo;
Sharpening picture obtains module, for carrying out Edge contrast to described gray scale picture, obtains sharpening picture;
Pixel value determination module, for determining the pixel value of each pixel of described sharpening picture;
Standard deviation determination module, for the pixel value of each pixel according to described sharpening picture, determines the standard deviation of the pixel value for described Target Photo, if described standard deviation is lower than the standard deviation threshold method preset, then triggers blurred picture determination module;
Described blurred picture determination module, for being defined as blurred picture by described Target Photo.
7. device according to claim 6, is characterized in that, described standard deviation determination module, comprising:
Judge submodule, for each pixel for described sharpening picture, judge that the pixel value of this pixel is whether in the numerical range preset, and if so, then filters out the pixel value of this pixel, if not, then retains the pixel value of this pixel;
Calculating sub module, for calculating the standard deviation of the pixel value of the pixel of reservation;
Standard deviation determination submodule, for being defined as the standard deviation of the pixel value for described Target Photo by the standard deviation calculated.
8. device according to claim 6, is characterized in that, described gray scale picture obtains module, specifically for:
According to the first proportionate relationship preset, process is reduced to described Target Photo, carrying out gray proces to reducing the picture after process, obtaining the gray scale picture that described Target Photo is corresponding;
Or,
Obtain and described Target Photo gray scale picture of the same size, according to the second proportionate relationship preset, process is reduced to obtained gray scale picture, obtains the gray scale picture that described Target Photo is corresponding.
9. device according to claim 6, is characterized in that, also comprises post-processing module:
Described post-processing module, for after described Target Photo is defined as blurred picture, exports the information of whether deleting described Target Photo, according to the selection of user for described information, determines whether to perform the operation of deleting described Target Photo;
Or,
The described Target Photo of direct deletion;
Or,
Described Target Photo is put into default picture file folder to be deleted.
10. the device according to any one of claim 6 to 9, is characterized in that, described sharpening picture obtains module, specifically for:
Use edge detection operator to carry out Edge contrast to described gray scale picture, obtain sharpening picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510438070.XA CN105069783B (en) | 2015-07-23 | 2015-07-23 | Fuzzy picture identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510438070.XA CN105069783B (en) | 2015-07-23 | 2015-07-23 | Fuzzy picture identification method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105069783A true CN105069783A (en) | 2015-11-18 |
| CN105069783B CN105069783B (en) | 2019-03-01 |
Family
ID=54499140
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510438070.XA Expired - Fee Related CN105069783B (en) | 2015-07-23 | 2015-07-23 | Fuzzy picture identification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105069783B (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512221A (en) * | 2015-11-30 | 2016-04-20 | 上海斐讯数据通信技术有限公司 | Fuzzy photo processing method and processing system |
| CN105808663A (en) * | 2016-02-29 | 2016-07-27 | 北京金山安全软件有限公司 | Image classification method and device and electronic equipment |
| CN106357982A (en) * | 2016-11-11 | 2017-01-25 | 北京佳艺徕经贸有限责任公司 | Image service device and method |
| CN106373121A (en) * | 2016-09-18 | 2017-02-01 | 广东小天才科技有限公司 | Fuzzy image identification method and apparatus |
| CN106407441A (en) * | 2016-09-28 | 2017-02-15 | 北京小米移动软件有限公司 | Mistaken photo identification method and device |
| CN106920266A (en) * | 2015-12-28 | 2017-07-04 | 腾讯科技（深圳）有限公司 | The Background Generation Method and device of identifying code |
| CN109214995A (en) * | 2018-08-20 | 2019-01-15 | 阿里巴巴集团控股有限公司 | The determination method, apparatus and server of picture quality |
| CN109829859A (en) * | 2018-12-05 | 2019-05-31 | 平安科技（深圳）有限公司 | Image processing method and terminal device |
| CN110458790A (en) * | 2018-05-03 | 2019-11-15 | 优酷网络技术（北京）有限公司 | A kind of image detecting method, device and computer storage medium |
| CN110704380A (en) * | 2019-08-27 | 2020-01-17 | 努比亚技术有限公司 | Automatic picture deleting method, terminal and computer readable storage medium |
| CN111738282A (en) * | 2019-10-22 | 2020-10-02 | 腾讯科技（深圳）有限公司 | Image recognition method based on artificial intelligence and related equipment |
| CN112396050A (en) * | 2020-12-02 | 2021-02-23 | 上海优扬新媒信息技术有限公司 | Image processing method, device and storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5166722A (en) * | 1990-08-14 | 1992-11-24 | Nikon Corporation | Camera image shake detecting apparatus |
| US20050243351A1 (en) * | 2004-04-19 | 2005-11-03 | Tatsuya Aoyama | Image processing method, apparatus, and program |
| CN102509285A (en) * | 2011-09-28 | 2012-06-20 | 宇龙计算机通信科技(深圳)有限公司 | Processing method and system for shot fuzzy picture and shooting equipment |
| CN102598687A (en) * | 2009-10-10 | 2012-07-18 | 汤姆森特许公司 | Method and device for calculating blur in video images |
| CN102665088A (en) * | 2010-12-22 | 2012-09-12 | 特克特朗尼克公司 | Blur detection with local sharpness map |
| CN103413311A (en) * | 2013-08-19 | 2013-11-27 | 厦门美图网科技有限公司 | Edge-based fuzzy detection method |
| CN104200480A (en) * | 2014-09-17 | 2014-12-10 | 西安电子科技大学宁波信息技术研究院 | Image fuzzy degree evaluation method and system applied to intelligent terminal |
| CN104268888A (en) * | 2014-10-09 | 2015-01-07 | 厦门美图之家科技有限公司 | Method for detecting image blurring |
-
2015
- 2015-07-23 CN CN201510438070.XA patent/CN105069783B/en not_active Expired - Fee Related
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5166722A (en) * | 1990-08-14 | 1992-11-24 | Nikon Corporation | Camera image shake detecting apparatus |
| US20050243351A1 (en) * | 2004-04-19 | 2005-11-03 | Tatsuya Aoyama | Image processing method, apparatus, and program |
| CN102598687A (en) * | 2009-10-10 | 2012-07-18 | 汤姆森特许公司 | Method and device for calculating blur in video images |
| CN102665088A (en) * | 2010-12-22 | 2012-09-12 | 特克特朗尼克公司 | Blur detection with local sharpness map |
| CN102509285A (en) * | 2011-09-28 | 2012-06-20 | 宇龙计算机通信科技(深圳)有限公司 | Processing method and system for shot fuzzy picture and shooting equipment |
| CN103413311A (en) * | 2013-08-19 | 2013-11-27 | 厦门美图网科技有限公司 | Edge-based fuzzy detection method |
| CN104200480A (en) * | 2014-09-17 | 2014-12-10 | 西安电子科技大学宁波信息技术研究院 | Image fuzzy degree evaluation method and system applied to intelligent terminal |
| CN104268888A (en) * | 2014-10-09 | 2015-01-07 | 厦门美图之家科技有限公司 | Method for detecting image blurring |
Non-Patent Citations (3)
| Title |
|---|
| CHUNG YUN-CHUNG 等: "A non-parametric blur measure based on edge analysis for image processing applications", 《IEEE CONFERENCE ON CYBERNETICS & INTELLIGENT SYSTEMS》 * |
| ELENA TSOMKO 等: "Efficient Method of Detecting Globally Blurry or Sharp Images", 《NINTH INTERNATIONAL WORKSHOP ON IMAGES ANALYSIS FOR MULTIMEDIA INTERACTIVE SERVICES》 * |
| 李鸿林 等: "无参考模糊图像质量评价改进算法", 《计算机应用》 * |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105512221A (en) * | 2015-11-30 | 2016-04-20 | 上海斐讯数据通信技术有限公司 | Fuzzy photo processing method and processing system |
| CN106920266B (en) * | 2015-12-28 | 2019-11-05 | 腾讯科技（深圳）有限公司 | The Background Generation Method and device of identifying code |
| CN106920266A (en) * | 2015-12-28 | 2017-07-04 | 腾讯科技（深圳）有限公司 | The Background Generation Method and device of identifying code |
| CN105808663A (en) * | 2016-02-29 | 2016-07-27 | 北京金山安全软件有限公司 | Image classification method and device and electronic equipment |
| CN105808663B (en) * | 2016-02-29 | 2019-12-03 | 北京金山安全软件有限公司 | Image classification method and device and electronic equipment |
| CN106373121A (en) * | 2016-09-18 | 2017-02-01 | 广东小天才科技有限公司 | Fuzzy image identification method and apparatus |
| CN106407441A (en) * | 2016-09-28 | 2017-02-15 | 北京小米移动软件有限公司 | Mistaken photo identification method and device |
| CN106357982A (en) * | 2016-11-11 | 2017-01-25 | 北京佳艺徕经贸有限责任公司 | Image service device and method |
| CN106357982B (en) * | 2016-11-11 | 2023-04-07 | 北京佳艺徕经贸有限责任公司 | Image service device and method |
| CN110458790A (en) * | 2018-05-03 | 2019-11-15 | 优酷网络技术（北京）有限公司 | A kind of image detecting method, device and computer storage medium |
| CN109214995A (en) * | 2018-08-20 | 2019-01-15 | 阿里巴巴集团控股有限公司 | The determination method, apparatus and server of picture quality |
| CN109829859A (en) * | 2018-12-05 | 2019-05-31 | 平安科技（深圳）有限公司 | Image processing method and terminal device |
| CN109829859B (en) * | 2018-12-05 | 2024-05-24 | 平安科技（深圳）有限公司 | Image processing method and terminal equipment |
| CN110704380A (en) * | 2019-08-27 | 2020-01-17 | 努比亚技术有限公司 | Automatic picture deleting method, terminal and computer readable storage medium |
| CN111738282A (en) * | 2019-10-22 | 2020-10-02 | 腾讯科技（深圳）有限公司 | Image recognition method based on artificial intelligence and related equipment |
| CN112396050A (en) * | 2020-12-02 | 2021-02-23 | 上海优扬新媒信息技术有限公司 | Image processing method, device and storage medium |
| CN112396050B (en) * | 2020-12-02 | 2023-09-15 | 度小满科技(北京)有限公司 | Image processing method, device and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105069783B (en) | 2019-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105069783A (en) | Fuzzy picture identification method and device | |
| US9349165B2 (en) | Automatically suggesting regions for blur kernel estimation | |
| US7860332B2 (en) | Method and system for determining an indication of focus of an image | |
| US9071745B2 (en) | Automatic capturing of documents having preliminarily specified geometric proportions | |
| US20190208119A1 (en) | Image processing method and apparatus | |
| US7305127B2 (en) | Detection and manipulation of shadows in an image or series of images | |
| US20200349716A1 (en) | Interactive image matting method, computer readable memory medium, and computer device | |
| US9466096B2 (en) | Deblurring images having spatially varying blur | |
| CN112714255B (en) | Shooting method and device, electronic equipment and readable storage medium | |
| US9894285B1 (en) | Real-time auto exposure adjustment of camera using contrast entropy | |
| US7869064B2 (en) | Printing device | |
| CN103548056A (en) | Image-processing device, image-capturing device, and image-processing method | |
| Brocher | Qualitative and quantitative evaluation of two new histogram limiting binarization algorithms | |
| JP5703255B2 (en) | Image processing apparatus, image processing method, and program | |
| US20150110403A1 (en) | Automatically determining the size of a blur kernel | |
| EP3035289A1 (en) | Method and device for adsorbing straight line/line segment, and method and device for constructing polygon | |
| CN103593120A (en) | Method and device for adhering screenshot box to boundary of region of interest during screenshot | |
| CN105827963A (en) | Scene changing detection method during shooting process and mobile terminal | |
| CN105719251B (en) | A kind of compression degraded image restored method that Linear Fuzzy is moved for big picture | |
| Kim et al. | A texture-aware salient edge model for image retargeting | |
| CN107886518A (en) | Picture detection method, device, electronic equipment and read/write memory medium | |
| US20240127404A1 (en) | Image content extraction method and apparatus, terminal, and storage medium | |
| CN110956584A (en) | Pathological microscopic image interactive display method, medium and equipment | |
| Yu et al. | A new dehazing algorithm based on overlapped sub-block homomorphic filtering | |
| CN105574823A (en) | Deblurring method and device for out-of-focus blurred image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190301 |