CN106127752A - Image quality analysis method and device - Google Patents
Image quality analysis method and device Download PDFInfo
- Publication number
- CN106127752A CN106127752A CN201610448288.8A CN201610448288A CN106127752A CN 106127752 A CN106127752 A CN 106127752A CN 201610448288 A CN201610448288 A CN 201610448288A CN 106127752 A CN106127752 A CN 106127752A
- Authority
- CN
- China
- Prior art keywords
- neighborhood
- picture
- picture quality
- point
- threshold value
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- 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
- Engineering & Computer Science (AREA)
- Quality & Reliability (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The disclosure is directed to a kind of image quality analysis method and device, belong to technical field of image processing.Wherein, method includes: acquisition processes the key point of picture；Key point is carried out clustering processing and generates multiple polymerization coordinate points；Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement；Picture quality according to neighborhood determines the picture quality of pending picture.The method increase accuracy rate and the efficiency of the picture quality determining pending picture.
Description
Technical field
It relates to technical field of image processing, particularly relate to a kind of image quality analysis method and device.
Background technology
Along with popularizing of the terminal units such as smart mobile phone, people get used to carrying out taking pictures to record life by terminal unit
Drop, and the picture of some user shooting is probably due to occur fuzzy, and causes the of low quality of picture, therefore, in order to save
Save the internal memory of terminal unit, can detect that picture of low quality, to carry out relevant treatment, such as carries out deletion etc..
In correlation technique, can use related algorithm that picture quality carries out correlation computations, wherein calculate picture quality
Input is whole image.
Summary of the invention
For overcoming problem present in correlation technique, disclosure embodiment provides a kind of image quality analysis method and dress
Put.Described technical scheme is as follows:
First aspect according to disclosure embodiment, it is provided that a kind of image quality analysis, the method includes:
Obtain the key point of pending picture；
Described key point is carried out clustering processing and generates multiple polymerization coordinate points；
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement；
Picture quality according to described neighborhood determines the picture quality of described pending picture.
Second aspect according to disclosure embodiment, it is provided that a kind of image quality analysis device, this device includes:
Extraction module, for obtaining the key point of pending picture；
Generation module, generates multiple polymerization coordinate points for described key point carries out clustering processing；
Judge module, for carrying out picture quality judgement to the neighborhood that multiple polymerization coordinate points are corresponding respectively；
Determine module, for determining the picture quality of described pending picture according to the picture quality of described neighborhood.
The third aspect according to disclosure embodiment, it is provided that another kind of image quality analysis device, described device includes:
Processor；
For storing the memorizer of the executable instruction of described processor；
Wherein, described processor is configured to:
Obtain the key point of pending picture；
Described key point is carried out clustering processing and generates multiple polymerization coordinate points；
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement；
Picture quality according to described neighborhood determines the picture quality of described pending picture.
The technical scheme that disclosure embodiment provides can include following beneficial effect:
Obtain the key point of pending picture, and multiple key points are carried out the clustering processing multiple polymerization coordinates of generation, and
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement, determine pending figure according to the picture quality of neighborhood
The picture quality of sheet.Thus, only key point place neighborhood is carried out picture quality judgement, improve the figure determining pending picture
The accuracy rate of picture element amount and efficiency.
It should be appreciated that it is only exemplary and explanatory, not that above general description and details hereinafter describe
The disclosure can be limited.
Accompanying drawing explanation
Accompanying drawing herein is merged in description and constitutes the part of this specification, it is shown that meet the enforcement of the disclosure
Example, and it is configured to explain the principle of the disclosure together with description.
Fig. 1 is the flow chart according to a kind of image quality analysis method shown in an exemplary embodiment；
Fig. 2 is according to the one shown in an exemplary embodiment, key point to be carried out clustering processing to generate multiple polymerization coordinates
The exemplary plot of point；
Fig. 3 is the flow chart according to a kind of image quality analysis method shown in another exemplary embodiment；
Fig. 4 is the figure determining pending picture according to a kind of picture quality according to neighborhood shown in an exemplary embodiment
The flow chart of the method for picture element amount；
Fig. 5 is the block diagram according to a kind of image quality analysis device shown in an exemplary embodiment；
Fig. 6 is the block diagram according to a kind of image quality analysis device shown in another exemplary embodiment；
Fig. 7 is the block diagram according to a kind of image quality analysis device shown in further example embodiment；And
Fig. 8 is the block diagram according to a kind of image quality analysis device shown in another exemplary embodiment.
By above-mentioned accompanying drawing, it has been shown that the embodiment that the disclosure is clear and definite, hereinafter will be described in more detail.These accompanying drawings
With word, the scope being not intended to be limited disclosure design by any mode is described, but by with reference to specific embodiment being
Those skilled in the art illustrate the concept of the disclosure.
Detailed description of the invention
Here will illustrate exemplary embodiment in detail, its example represents in the accompanying drawings.Explained below relates to
During accompanying drawing, unless otherwise indicated, the same numbers in different accompanying drawings represents same or analogous key element.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistent with the disclosure.On the contrary, they are only with the most appended
The example of the apparatus and method that some aspects that described in detail in claims, the disclosure are consistent.
Below with reference to the accompanying drawings image quality analysis method and the device of disclosure embodiment are described.
Fig. 1 is the flow chart according to a kind of image quality analysis method shown in an exemplary embodiment.
As it is shown in figure 1, this image quality analysis method can include following several step:
In step s 110, the key point of pending picture is obtained.
In actual applications, user is when taking pictures, in order to highlight its target object taken pictures, and target that can be to be taken pictures to it
Object is deliberately focused and is obtained background blurring photo, therefore, if the simple overall picture quality judging image, if
When the background area of virtualization process is bigger, it is likely that the background region by blurring is affected, and is judged as image
Picture of low quality.
Therefore, in one embodiment, in order to accurately judge picture quality, it is only necessary to judge the figure of target object region
Picture element amount.
Specifically, the image quality analysis method of disclosure embodiment, utilize key point extraction algorithm, such as utilize
SIFI, Harris algorithms etc. carry out gradient and extract the key point of pending picture pending picture, and wherein, key point corresponds to
Some flex points that in image, gradient is bigger.
And due in picture the key point of background region of virtualization considerably less or almost without, therefore this key point
Place neighborhood is the region that user wishes the target object place of shooting, thus only by the image to key point neighbors around
Quality estimation can determine the picture quality of pending picture accurately.
Wherein, in an embodiment of the disclosure, in order to be further ensured that the key point of extraction corresponding be Grad
Bigger neighborhood, avoids the background region of the virtualization impact on picture quality when judging picture quality further, also may be used
One Grads threshold is set.
The most only when Grad is more than this Grads threshold, just extract it for key point.Such as, can calculate according to Corner Detection
Method is retrieved as key point by extracting in pending picture more than the point of the Grads threshold preset.
In the step s 120, key point is carried out clustering processing and generates multiple polymerization coordinate points.
In embodiment of the disclosure, for the ease of computational analysis, by the most scattered pass of the pending picture of acquisition
Key point carries out clustering processing, and key point that will be the most scattered is clustered into multiclass according to its place neighborhood, corresponding one of each class
Polymerization coordinate points, this polymerization coordinate points may correspond to the center of the key point place neighborhood of cluster.
For example, extract to obtain key point as in figure 2 it is shown, pending picture to be carried out gradient, closed in a large number
After key point, key point is clustered, cluster as n neighborhood, and generate central point i.e. n the polymerization coordinate points of this neighborhood, point
Wei (x1, y1)-(xn, yn) (only marking the polymerization coordinate points of neighborhood 1 and neighborhood 1 in figure).
Wherein it is desired to explanation, according to the difference of application scenarios, can use multiple clustering algorithm that key point is gathered
Class processes.Such as, the clustering algorithm such as k-means algorithm, SOM clustering algorithm can be used to carry out clustering processing, by the most scattered
Key point is clustered into multiclass according to its place neighborhood, and the corresponding polymerization coordinate points of each class, disclosure embodiment is to employing
Aggregating algorithm is not construed as limiting.
In step s 130, respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement.
Illustratively, corresponding to the multiple polymerization coordinate points respectively each neighborhood in multiple neighborhoods carries out picture quality and sentences
Disconnected, if multiple neighborhoods of correspondence are n, then carry out the judgement of n picture quality.
It should be noted that according to the difference of concrete application scenarios, multiple determination methods can be used to judge each polymerization coordinate
The picture quality of the neighborhood that point is corresponding.
The first example, can calculate the quantity of dead pixel points of images in the neighborhood that each polymerization coordinate points is corresponding respectively, work as figure
When being more than certain value as the quantity of bad point, it is judged that this neighborhood is the neighborhood that picture quality is fuzzy.In the present embodiment, dead pixel points of images
Referring to compared with surrounding pixel point, the difference of colour and described surrounding pixel point is more than the pixel of preset value.This preset value example
As obtained according to the experience of those skilled in the art or test of many times, this is not construed as limiting by disclosure embodiment.
The second example, can calculate the Grad in the neighborhood that each polymerization coordinate points is corresponding respectively, when many in this neighborhood
When the Grad of individual pixel is more than certain value, it is determined that the good image quality of this neighborhood.
In step S140, determine the picture quality of pending picture according to the picture quality of neighborhood.
Specifically, owing to above-mentioned each neighborhood is the neighborhood generated after key point clusters, the therefore picture quality of this each neighborhood
The picture quality of pending picture can be characterized.
Wherein it is desired to explanation, determine that according to the picture quality of neighborhood the mode of the picture quality of pending picture has
Multiple.
The first example, can arrange the mark of correspondence respectively, be in multiple neighborhood edge according to each neighborhood position
The mark that arranges of neighborhood minimum, be in the mark that the neighborhood of multiple centre of neighbourhood position arranges the highest, thus respectively according to many
The picture quality of individual neighborhood is given a mark.
In this example, preset a mark S, and if then the picture quality of current neighborhood clear, then to S plus corresponding
Mark, if the picture quality of current neighborhood obscures, then deducts corresponding mark to S, the like, until multiple neighborhoods are beaten
Divide complete, if final score S is more than certain value, it is determined that pending picture quality is clear, if S is less than or equal to certain value,
Then determine that the picture quality of pending picture obscures.
The second example, clearly can determine the image of pending picture with fuzzy quantity according to picture quality in neighborhood
Quality.
If the good image quality of major part neighborhood in the neighborhood that i.e. each polymerization coordinate points is corresponding, then it is assumed that this is pending
The good image quality of picture；If the picture quality of major part neighborhood obscures in the neighborhood that respectively polymerization coordinate points is corresponding, then recognize
Picture quality for this pending picture obscures.
In sum, the image quality analysis method of disclosure embodiment, obtain the key point of pending picture, and to many
Individual key point carries out clustering processing and generates multiple polymerization coordinates, and respectively the neighborhood that each polymerization coordinate points is corresponding is carried out image
Quality estimation, determines the picture quality of pending picture according to the picture quality of neighborhood.Thus, only to key point place neighborhood
Carry out picture quality judgement, improve accuracy rate and the efficiency of the picture quality determining pending picture.
Based on above-described embodiment, in order in more detailed explanation above-mentioned steps S130, how to multiple polymerization coordinate points
Corresponding each neighborhood carries out picture quality judgement, and below in conjunction with the accompanying drawings 3, with the Grad of the image by calculating each neighborhood, point
Other the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement as a example by illustrate, be described as follows:
In step S310, obtain the key point of pending picture.
In step s 320, key point is carried out clustering processing and generates multiple polymerization coordinate points.
It should be noted that step S310-S320 is corresponding with above-mentioned steps S110-S120, therefore to step S310-
S320 describes with reference to the description to step S110-S120, does not repeats them here.
In step S330, respectively the neighborhood that each polymerization coordinate points is corresponding is carried out gradient detection.
In step S340, detect the pixel neighborhood of a point that whether there is quantity in each neighborhood more than predetermined threshold value, should
The Grad of pixel exceedes predetermined gradient value.S350, when there is the pixel neighborhood of a point that quantity is more than predetermined threshold value, really
Determined number is clear more than the picture quality of the pixel neighborhood of a point of predetermined threshold value.
S360, when there is not the pixel neighborhood of a point that quantity is more than predetermined threshold value, determining and there is not quantity more than pre-
If the picture quality of the pixel neighborhood of a point of threshold value obscures.
Illustratively, respectively the neighborhood that each polymerization coordinate points is corresponding is carried out the calculating of Grad, detects in each neighborhood
Whether Grad is more than predetermined gradient value, and wherein, predetermined gradient value is demarcated according to great many of experiments.
If the Grad of the current pixel point calculated is more than predetermined gradient value, then it is assumed that the pixel value of this current pixel point
Higher, picture quality corresponding to current pixel point is clear, if the Grad of the current pixel point calculated is not more than predetermined gradient
Value, then it is assumed that the pixel value of this current pixel point is relatively low, picture quality corresponding to current pixel point obscures.
In embodiment of the disclosure, when respectively the neighborhood that each polymerization coordinate points is corresponding being carried out gradient detection, based on
The picture quality of the size estimation current neighborhood of Grad, it is not intended that the direction of Grad, is therefore sitting each polymerization respectively
When the neighborhood that punctuate is corresponding carries out gradient detection, the absolute value obtaining Grad judges the picture quality of current neighborhood.
For example, when the neighborhood that size is 25*25 that polymerization coordinate points (x1, y1) is corresponding is carried out gradient detection, as
Really the first behavior 5 100 in this neighborhood, the second behavior 5 10, then the Grad obtained after subtracting each other is 5 90；If this neighborhood
Interior first behavior 5 10, the second behavior 5 100, is 5-90 after subtracting each other, and the absolute value the most now taking Grad obtains 5
90。
If it is further appreciated that the Grad of the interior only a few pixels point of current neighborhood is relatively low, and major part ladder
Angle value is higher, then it is assumed that the picture quality of this current neighborhood is clear.Only at the relatively low pixel of current neighborhood manhole ladder angle value
When quantity is more than some, just can determine that the picture quality of current neighborhood obscures.
Illustratively, can arrange according to the size of lot of experimental data and current neighborhood with predetermined threshold value, if worked as
Front neighborhood exceedes the pixel quantity of predetermined gradient value more than predetermined threshold value, it is determined that the picture quality of current neighborhood is clear
Clear.
If current neighborhood exceeding the pixel quantity of predetermined gradient value less than or equal to predetermined threshold value, it is determined that current
The picture quality of neighborhood obscures.
In step S370, determine the picture quality of pending picture according to the picture quality of neighborhood.
In sum, the image quality analysis method of disclosure embodiment, by calculating the Grad of the image of each neighborhood,
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement, and then determine pending according to the picture quality of neighborhood
The picture quality of picture.Thus, improve accuracy rate and the efficiency of the picture quality determining pending picture.
Based on above-described embodiment, further, in order to describe in detail in disclosure embodiment, how according to the image of neighborhood
Quality determines the picture quality of pending picture, below in conjunction with the accompanying drawings 4, and with clear with fuzzy according to picture quality in each neighborhood
Quantity illustrates as a example by determining the picture quality of pending picture, is described as follows:
Fig. 4 is the figure determining pending picture according to a kind of picture quality according to neighborhood shown in an exemplary embodiment
The flow chart of the method for picture element amount, as shown in Figure 4, the method includes:
In step S410, in detection neighborhood, whether picture quality neighborhood quantity clearly is more than predetermined threshold value.
In the step s 420, when picture quality neighborhood quantity clearly is more than predetermined threshold value, pending picture is determined
Picture quality is clear.
In step S430, in picture quality neighborhood quantity clearly less than or equal to predetermined threshold value, determine pending picture
Picture quality obscure.
It is appreciated that previously according to the size of each neighborhood in lot of experimental data, multiple neighborhood and pending picture
Size predetermined threshold value is set, if the size of each neighborhood of multiple neighborhood is less, pending picture is relatively big, then this predetermined threshold value
Bigger；If each neighborhood of multiple neighborhoods is relatively large, pending picture is less, then this predetermined threshold value is less.
And then, detect whether picture quality neighborhood quantity clearly in multiple neighborhood is more than predetermined threshold value, if figure picture element
Measure neighborhood quantity clearly and be more than predetermined threshold value, it is determined that the picture quality of pending picture is clear.
If picture quality neighborhood quantity clearly is less than or equal to predetermined threshold value, it is determined that the picture quality of pending picture
Fuzzy.
In sum, the image quality analysis method of disclosure embodiment, according to picture quality in multiple neighborhoods clear with
Fuzzy neighborhood quantity determines the picture quality of pending picture, when the picture quality of major part neighborhood is clear, it is determined that treat
The picture quality processing picture is clear, when the picture quality of only sub-fraction neighborhood is clear, it is determined that pending picture
Picture quality obscures, and which thereby enhances accuracy rate and the efficiency of the picture quality determining pending picture.
Disclosure image quality analysis device embodiment is described below in detail, may be used for performing disclosure graphical analysis
Embodiment of the method.For details the most disclosed in disclosure image quality analysis device embodiment, refer to disclosure image
Analyze embodiment of the method.
Fig. 5 is the block diagram according to a kind of image quality analysis device shown in an exemplary embodiment, as it is shown in figure 5, should
Image quality analysis device includes:
Extraction module 510, is configured to obtain the key point of pending picture.
Generation module 520, is configured to that key point carries out clustering processing and generates multiple polymerization coordinate points.
Judge module 530, is configured to the neighborhood to each polymerization coordinate points is corresponding and carries out picture quality judgement.
Determine module 540, be configured to the picture quality according to neighborhood and determine the picture quality of pending picture.
Specifically, the image quality analysis device of disclosure embodiment, extraction module 510 utilizes key point extraction algorithm,
Such as extraction module 510 utilizes SIFI, Harris algorithm etc. that pending picture is carried out gradient and extracts the key of pending picture
Point, wherein, key point corresponds to some flex points that in image, gradient is bigger.
And due in picture the key point of background place neighborhood of virtualization considerably less or almost without, therefore this key point
Place neighborhood is the region that user wishes the target object place of shooting, thus only by the picture quality to key point neighborhood
Judgement can determine the picture quality of pending picture accurately.
Wherein, in an embodiment of the disclosure, in order to be further ensured that the key point of extraction corresponding be Grad
Bigger neighborhood, avoids the background place neighborhood of the virtualization impact on picture quality when judging picture quality further, also may be used
One Grads threshold is set.
I.e. extraction module 510 is only when Grad is more than this Grads threshold, just extracts it for key point.Such as, can root
It is retrieved as key point by pending picture extracts more than the point of the Grads threshold preset according to Corner Detection Algorithm.
Further, for the ease of computational analysis, generation module 520 is by the most scattered pass of the pending picture of acquisition
Key point carries out clustering processing, and key point that will be the most scattered is clustered into multiclass according to its place neighborhood, corresponding one of each class
Polymerization coordinate points, this polymerization coordinate points may correspond to the center of the key point place neighborhood of cluster.
And then, it is judged that each neighborhood in the neighborhood that module 530 is corresponding to each polymerization coordinate points respectively carries out figure picture element
Amount judges, if multiple neighborhoods of correspondence are n, then carries out the judgement of n picture quality.Further, due to above-mentioned respectively
Neighborhood is the multiple neighborhoods generated after key point clusters, and therefore the picture quality of this each neighborhood can characterize the figure of pending picture
Picture element amount, can be determined by module 540 and determine the picture quality of pending picture according to the picture quality of multiple neighborhoods.
In sum, the image quality analysis device of disclosure embodiment, obtain the key point of pending picture, and to many
Individual key point carries out clustering processing and generates multiple polymerization coordinates, and respectively the neighborhood that each polymerization coordinate points is corresponding is carried out figure picture element
Amount judges, determines the picture quality of pending picture according to the picture quality of neighborhood.Thus, only key point place neighborhood is entered
Row picture quality judges, improves accuracy rate and the efficiency of the picture quality determining pending picture.
Based on above-described embodiment, in order to the more detailed multiple neighborhoods illustrated how each polymerization coordinate points is corresponding are carried out
Picture quality judges, below in conjunction with the accompanying drawings 6, and with the Grad of the image by calculating multiple neighborhoods, respectively to each polymerization coordinate
The neighborhood that point is corresponding illustrates as a example by carrying out picture quality judgement, is described as follows:
Fig. 6 is the block diagram according to a kind of image quality analysis device shown in another exemplary embodiment, as shown in Figure 6,
As shown in Figure 6, on the basis of as shown in Figure 5, it is judged that module 530 includes:
First detector unit 531, is configured to the neighborhood to each polymerization coordinate points is corresponding and carries out gradient detection.
Second detector unit 532, is configured to detect whether each neighborhood exists the quantity pixel more than predetermined threshold value
Neighborhood, the Grad of this pixel exceedes predetermined gradient value.First determines unit 533, is configured to there is quantity more than pre-
If during the pixel neighborhood of a point of threshold value, quantification is clear more than the picture quality of the pixel neighborhood of a point of predetermined threshold value.
Specifically, the first detector unit 531 carries out the calculating of Grad respectively to the neighborhood that each polymerization coordinate points is corresponding, the
Two detector units 532 detect whether the Grad in each neighborhood is more than predetermined gradient value, and wherein, predetermined gradient value is according to big
Amount experimental calibration.
If the Grad of the current pixel point calculated is more than predetermined gradient value, then it is assumed that the pixel value of this current pixel point
Higher, picture quality corresponding to current pixel point is clear, if the Grad of the current pixel point calculated is not more than predetermined gradient
Value, then it is assumed that the pixel value of this current pixel point is relatively low, picture quality corresponding to current pixel point obscures.
Wherein, in embodiment of the disclosure, respectively multiple neighborhoods that polymerization coordinate points is corresponding are being carried out gradient detection
Time, the picture quality of size estimation current neighborhood based on Grad, it is not intended that the direction of Grad, therefore respectively to respectively
When polymerization neighborhood corresponding to coordinate points carries out gradient detection, the second detector unit 532 obtains the absolute value of Grad and judges current
The picture quality of neighborhood.
And then, it will be understood that if the Grad of the interior only a few pixels point of current neighborhood is relatively low, and major part Grad
Higher, then it is assumed that the picture quality of this current neighborhood is clear.Only in the quantity of the relatively low pixel of current neighborhood manhole ladder angle value
During more than some, just can determine that the picture quality of current neighborhood obscures.
Specifically, arrange according to the size of lot of experimental data and current neighborhood with predetermined threshold value, if when current
The quantity exceeding predetermined gradient value in neighborhood is more than predetermined threshold value, and first determines that unit 533 then determines the image of current neighborhood
Quality is clear.
If the quantity exceeding predetermined gradient value in current neighborhood is less than or equal to predetermined threshold value, first determines unit 533
Then determine that the picture quality of current neighborhood obscures.
In sum, the image quality analysis device of disclosure embodiment, by calculating the Grad of the image of each neighborhood,
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement, and then determine pending figure according to multiple picture qualities
The picture quality of sheet.Thus, improve accuracy rate and the efficiency of the picture quality determining pending picture.
Based on above-described embodiment, further, in order to describe in detail in disclosure embodiment, how according to the image of neighborhood
Quality determines the picture quality of pending picture, below in conjunction with the accompanying drawings 7, and with the neighbour clear and fuzzy according to picture quality in neighborhood
Territory quantity illustrates as a example by determining the picture quality of pending picture, is described as follows:
Fig. 7 is the block diagram according to a kind of image quality analysis device shown in further example embodiment, as it is shown in fig. 7,
On the basis of as shown in Figure 5, determine that module 540 includes:
3rd detector unit 541, is configured to detect whether picture quality neighborhood quantity clearly in neighborhood is more than default
Threshold value.
Second determines unit 542, is configured to, when picture quality neighborhood quantity clearly is more than predetermined threshold value, determine and treat
The picture quality processing picture is clear.
It is appreciated that previously according to the size of each neighborhood in lot of experimental data, many neighborhoods and pending picture
Size arranges predetermined threshold value, if the size of each neighborhood of multiple neighborhood is less, pending picture is relatively big, then this predetermined threshold value is relatively
Greatly；If each neighborhood of multiple neighborhoods is relatively large, pending picture is less, then this predetermined threshold value is less.
And then, the 3rd detector unit 541 detects whether picture quality neighborhood quantity clearly in each neighborhood is more than default threshold
Value, if picture quality neighborhood quantity clearly is more than predetermined threshold value, second determines that unit 542 then determines the figure of pending picture
Picture element amount is clear.
If picture quality neighborhood quantity clearly is less than or equal to predetermined threshold value, second determines that unit 542 then determines waits to locate
The picture quality of reason picture obscures.
In sum, the image quality analysis device of disclosure embodiment, clear and fuzzy according to picture quality in neighborhood
Neighborhood quantity determine the picture quality of pending picture, when the picture quality of major part neighborhood is clear, it is determined that pending
The picture quality of picture is clear, when the picture quality of only sub-fraction neighborhood is clear, it is determined that the image of pending picture
Quality obscures, and which thereby enhances accuracy rate and the efficiency of the picture quality determining pending picture.
Fig. 8 is the block diagram according to a kind of image quality analysis device shown in another exemplary embodiment.Such as, figure picture element
Component analysis device 1000 can be mobile phone, computer, digital broadcast terminal, messaging devices, game console, flat board
Equipment, armarium, body-building equipment, personal digital assistant etc..
With reference to Fig. 8, image quality analysis device 1000 can include following one or more assembly: processes assembly 1002,
Memorizer 1004, power supply module 1006, multimedia groupware 1008, audio-frequency assembly 1010, the interface 1012 of input/output (I/O),
Sensor cluster 1014, and communications component 1016.
Processing assembly 1002 and generally control the integrated operation of image quality analysis device 1000, such as with display, phone is exhaled
Cry, data communication, the operation that camera operation and record operation are associated.Process assembly 1002 and can include one or more process
Device 1020 performs instruction, to complete all or part of step of above-mentioned method.Additionally, process assembly 1002 can include one
Individual or multiple modules, it is simple to process between assembly 1002 and other assemblies is mutual.Such as, process assembly 1002 and can include many
Media module, with facilitate multimedia groupware 1008 and process between assembly 1002 mutual.
Memorizer 1004 is configured to store various types of data to support the behaviour at image quality analysis device 1000
Make.The example of these data includes any application program or the method being configured on image quality analysis device 1000 operation
Instruction, contact data, telephone book data, message, picture, video etc..Memorizer 1004 can be by any kind of volatile
Property or non-volatile memory device or combinations thereof realize, such as static RAM (SRAM), electric erasable can
Program read-only memory (EEPROM), Erasable Programmable Read Only Memory EPROM (EPROM), programmable read only memory (PROM),
Read only memory (ROM), magnetic memory, flash memory, disk or CD.
The various assemblies that power supply module 1006 is image quality analysis device 1000 provide electric power.Power supply module 1006 is permissible
Including power-supply management system, one or more power supplys, and other generate with for image quality analysis device 1000, manage and distribute
The assembly that electric power is associated.
Offer one output that multimedia groupware 1008 is included between described image quality analysis device 1000 and user
The touching display screen of interface.In certain embodiments, touching display screen can include liquid crystal display (LCD) and touch panel
(TP).Touch panel includes that one or more touch sensor is with the gesture on sensing touch, slip and touch panel.Described tactile
Touch sensor and can not only sense touch or the border of sliding action, but also detection is relevant to described touch or slide
Persistent period and pressure.In certain embodiments, multimedia groupware 1008 includes a front-facing camera and/or rearmounted shooting
Head.When image quality analysis device 1000 is in operator scheme, during such as screening-mode or video mode, front-facing camera and/or
Post-positioned pick-up head can receive the multi-medium data of outside.Each front-facing camera and post-positioned pick-up head can be one fixing
Optical lens system or there is focal length and optical zoom ability.
Audio-frequency assembly 1010 is configured to output and/or input audio signal.Such as, audio-frequency assembly 1010 includes a wheat
Gram wind (MIC), when image quality analysis device 1000 is in operator scheme, such as call model, logging mode and speech recognition mould
During formula, mike is configured to receive external audio signal.The audio signal received can be further stored at memorizer
1004 or send via communications component 1016.In certain embodiments, audio-frequency assembly 1010 also includes a speaker, is configured
For output audio signal.
I/O interface 1012 provides interface, above-mentioned peripheral interface module for processing between assembly 1002 and peripheral interface module
Can be keyboard, put striking wheel, button etc..These buttons may include but be not limited to: home button, volume button, start button and
Locking press button.
Sensor cluster 1014 includes one or more sensor, and being configured to image quality analysis device 1000 provides each
The state estimation of individual aspect.Such as, what sensor cluster 1014 can detect image quality analysis device 1000 beats opening/closing
State, the relative localization of assembly, the most described assembly is display and the keypad of image quality analysis device 1000, sensor
Assembly 1014 can also detect the position of image quality analysis device 1000 or 1,000 1 assemblies of image quality analysis device and change
Become, the presence or absence that user contacts with image quality analysis device 1000, image quality analysis device 1000 orientation or add
Speed/slow down and the variations in temperature of image quality analysis device 1000.Sensor cluster 1014 can include proximity transducer, is joined
Put object near the detection when not having any physical contact.Sensor cluster 1014 can also include that light senses
Device, such as CMOS or ccd image sensor, is configured in imaging applications use.In certain embodiments, this sensor cluster
1014 can also include acceleration transducer, gyro sensor, Magnetic Sensor, pressure transducer or temperature sensor.
Communications component 1016 is configured to facilitate between image quality analysis device 1000 and other equipment wired or wireless
The communication of mode.Image quality analysis device 1000 can access wireless network based on communication standard, such as WiFi, 2G or 3G,
Or combinations thereof.In one exemplary embodiment, communications component 1016 receives from external broadcasting pipe via broadcast channel
The broadcast singal of reason system or broadcast related information.In one exemplary embodiment, described communications component 1016 also includes closely
Field communication (NFC) module, to promote junction service.Such as, can be based on RF identification (RFID) technology, infrared number in NFC module
According to association (IrDA) technology, ultra broadband (UWB) technology, bluetooth (BT) technology and other technologies realize.
In the exemplary embodiment, image quality analysis device 1000 can be by one or more application specific integrated circuits
(ASIC), digital signal processor (DSP), digital signal processing appts (DSPD), PLD (PLD), scene can
Programming gate array (FPGA), controller, microcontroller, microprocessor or other electronic components realize, and are configured to perform above-mentioned
Image quality analysis method (method shown in Fig. 1-Fig. 4).
In the exemplary embodiment, a kind of non-transitory computer-readable recording medium including instruction, example are additionally provided
As included the memorizer 1004 of instruction, above-mentioned instruction can have been performed by the processor 1020 of image quality analysis device 1000
Said method.Such as, described non-transitory computer-readable recording medium can be ROM, random access memory (RAM), CD-
ROM, tape, floppy disk and optical data storage devices etc..
A kind of non-transitory computer-readable recording medium, when the instruction in described storage medium is held by the processor of terminal
During row so that terminal is able to carry out a kind of image quality analysis method, and described method includes:
Obtain the key point of pending picture；
Described key point is carried out clustering processing and generates multiple polymerization coordinate points；
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement；
Picture quality according to described neighborhood determines the picture quality of described pending picture.
Those skilled in the art, after considering description and putting into practice invention disclosed herein, will readily occur to its of the disclosure
Its embodiment.The disclosure is intended to any modification, purposes or the adaptations of the disclosure, these modification, purposes or
Person's adaptations is followed the general principle of the disclosure and includes the undocumented common knowledge in the art of the disclosure
Or conventional techniques means.Description and embodiments is considered only as exemplary, and the true scope of the disclosure and spirit are by following
Claim is pointed out.
It should be appreciated that the disclosure is not limited to precision architecture described above and illustrated in the accompanying drawings, and
And various modifications and changes can carried out without departing from the scope.The scope of the present disclosure is only limited by appended claim.
Claims (11)
1. an image quality analysis method, it is characterised in that comprise the following steps:
Obtain the key point of pending picture；
Described key point is carried out clustering processing and generates multiple polymerization coordinate points；
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement；
Picture quality according to described neighborhood determines the picture quality of described pending picture.
2. the method for claim 1, it is characterised in that the key point of the pending picture of described acquisition, including:
According to Corner Detection Algorithm, described pending picture will be retrieved as described key more than the point of the Grads threshold preset
Point.
3. the method for claim 1, it is characterised in that the described clustering processing that carries out described key point generates multiple poly-
Close coordinate points, including:
By clustering algorithm, described key point is gathered for multiple classes；
Key point Coordinate generation according to all kinds of correspondences and all kinds of corresponding polymerization coordinate points.
4. the method for claim 1, it is characterised in that described respectively the neighborhood that each polymerization coordinate points is corresponding is carried out figure
Picture Quality estimation, including:
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out gradient detection；
Detecting the pixel neighborhood of a point that whether there is quantity in each neighborhood more than predetermined threshold value, the Grad of described pixel surpasses
Cross predetermined gradient value；
When there is the pixel neighborhood of a point that described quantity is more than predetermined threshold value, determine that described quantity is more than predetermined threshold value
The picture quality of pixel neighborhood of a point is clear；
When there is not described quantity more than the pixel neighborhood of a point of predetermined threshold value, determine described in there is not quantity more than presetting
The picture quality of the pixel neighborhood of a point of threshold value obscures.
5. the method as described in any one of claim 1-4, it is characterised in that the described picture quality according to described neighborhood determines
The picture quality of described pending picture, including:
Detect whether picture quality neighborhood quantity clearly in described neighborhood is more than predetermined threshold value；
When picture quality neighborhood quantity clearly is more than predetermined threshold value, determine that the picture quality of described pending picture is clear；
When picture quality neighborhood quantity clearly is less than or equal to predetermined threshold value, determine the picture quality mould of described pending picture
Stick with paste.
6. an image quality analysis device, it is characterised in that including:
Extraction module, for obtaining the key point of pending picture；
Generation module, generates multiple polymerization coordinate points for described key point carries out clustering processing；
Judge module, for carrying out picture quality judgement to the neighborhood that each polymerization coordinate points is corresponding respectively；
Determine module, for determining the picture quality of described pending picture according to the picture quality of described neighborhood.
7. device as claimed in claim 6, it is characterised in that described extraction module is used for:
According to Corner Detection Algorithm, described pending picture will be retrieved as key point more than the point of the Grads threshold preset.
8. device as claimed in claim 6, it is characterised in that described generation module is used for:
By clustering algorithm, described key point is gathered for multiple classes；
Key point Coordinate generation according to all kinds of correspondences and all kinds of corresponding polymerization coordinate points.
9. device as claimed in claim 6, it is characterised in that described judge module includes:
First detector unit, for carrying out gradient detection to the neighborhood that each polymerization coordinate points is corresponding respectively；
Second detector unit, for detecting whether each neighborhood exists the quantity pixel neighborhood of a point more than predetermined threshold value, described
The Grad of pixel exceedes predetermined gradient value；
First determines unit, during for there is described quantity more than the pixel neighborhood of a point of predetermined threshold value, determines described quantity
Clear more than the picture quality of the pixel neighborhood of a point of predetermined threshold value；
Described first determines that unit is additionally operable to, when there is not the pixel neighborhood of a point that described quantity is more than predetermined threshold value, determine
The described quantity that do not exists obscures more than the picture quality of the pixel neighborhood of a point of predetermined threshold value.
10. device as claimed in claim 6, it is characterised in that described determine that module includes:
3rd detector unit, is used for detecting whether picture quality neighborhood quantity clearly in described neighborhood is more than predetermined threshold value；
Second determines unit, for when picture quality neighborhood quantity clearly is more than predetermined threshold value, determining described pending figure
The picture quality of sheet is clear；
Described second determines unit, is additionally operable to, when picture quality neighborhood quantity clearly is less than or equal to predetermined threshold value, determine institute
The picture quality stating pending picture obscures.
11. 1 kinds of image quality analysis devices, it is characterised in that described device includes:
Processor；
For storing the memorizer of the executable instruction of described processor；
Wherein, described processor is configured to:
Acquisition processes the key point of picture；
Described key point is carried out clustering processing and generates multiple polymerization coordinate points；
Respectively the neighborhood that each polymerization coordinate points is corresponding is carried out picture quality judgement；
Picture quality according to described neighborhood determines the picture quality of described pending picture.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610448288.8A CN106127752A (en) | 2016-06-20 | 2016-06-20 | Image quality analysis method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610448288.8A CN106127752A (en) | 2016-06-20 | 2016-06-20 | Image quality analysis method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106127752A true CN106127752A (en) | 2016-11-16 |
Family
ID=57470986
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610448288.8A Pending CN106127752A (en) | 2016-06-20 | 2016-06-20 | Image quality analysis method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106127752A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537787A (en) * | 2018-03-30 | 2018-09-14 | 中国科学院半导体研究所 | A kind of quality judging method of facial image |
| CN112446849A (en) * | 2019-08-13 | 2021-03-05 | 杭州海康威视数字技术股份有限公司 | Method and device for processing picture |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102176758A (en) * | 2011-03-07 | 2011-09-07 | 北京文安科技发展有限公司 | Video quality diagnosis system and realization method thereof |
| CN103020947A (en) * | 2011-09-23 | 2013-04-03 | 阿里巴巴集团控股有限公司 | Image quality analysis method and device |
| CN103632369A (en) * | 2013-12-05 | 2014-03-12 | 淮海工学院 | Method for universally detecting quality of non-reference underwater images on basis of combination of block average definition |
| US20140270464A1 (en) * | 2013-03-15 | 2014-09-18 | Mitek Systems, Inc. | Systems and methods for assessing standards for mobile image quality |
-
2016
- 2016-06-20 CN CN201610448288.8A patent/CN106127752A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102176758A (en) * | 2011-03-07 | 2011-09-07 | 北京文安科技发展有限公司 | Video quality diagnosis system and realization method thereof |
| CN103020947A (en) * | 2011-09-23 | 2013-04-03 | 阿里巴巴集团控股有限公司 | Image quality analysis method and device |
| CN103020947B (en) * | 2011-09-23 | 2016-04-06 | 阿里巴巴集团控股有限公司 | A kind of mass analysis method of image and device |
| US20140270464A1 (en) * | 2013-03-15 | 2014-09-18 | Mitek Systems, Inc. | Systems and methods for assessing standards for mobile image quality |
| CN103632369A (en) * | 2013-12-05 | 2014-03-12 | 淮海工学院 | Method for universally detecting quality of non-reference underwater images on basis of combination of block average definition |
Non-Patent Citations (3)
| Title |
|---|
| WUFENG XUE, LEI ZHANG, XUANQIN MOU: "Learning without Human Scores for Blind Image Quality Assessment", 《CVPR2013》 * |
| 李雷达: "《图像质量评价中的特征提取方法与应用》", 30 June 2015, 中国矿业大学出版社 * |
| 田小林，李骁一，杜文亮: "《光学相干层析图像处理及应用》", 31 January 2015, 北京理工大学出版社 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108537787A (en) * | 2018-03-30 | 2018-09-14 | 中国科学院半导体研究所 | A kind of quality judging method of facial image |
| CN112446849A (en) * | 2019-08-13 | 2021-03-05 | 杭州海康威视数字技术股份有限公司 | Method and device for processing picture |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105512685B (en) | Object identification method and device | |
| CN105809704B (en) | Identify the method and device of image definition | |
| CN104850828B (en) | Character recognition method and device | |
| CN109446994B (en) | Gesture key point detection method and device, electronic equipment and storage medium | |
| CN105488511B (en) | The recognition methods of image and device | |
| JP6335289B2 (en) | Method and apparatus for generating an image filter | |
| CN106572299A (en) | Camera switching-on method and device | |
| CN108668080B (en) | Method and device for prompting degree of dirt of lens and electronic equipment | |
| CN105069083B (en) | The determination method and device of association user | |
| CN105426867A (en) | Face identification verification method and apparatus | |
| CN110650379B (en) | Video abstract generation method and device, electronic equipment and storage medium | |
| CN105095881A (en) | Method, apparatus and terminal for face identification | |
| CN105528078B (en) | The method and device of controlling electronic devices | |
| CN104503888A (en) | Warning method and device | |
| CN104408404A (en) | Face identification method and apparatus | |
| CN105046231A (en) | Face detection method and device | |
| CN104284240A (en) | Video browsing method and device | |
| CN105631803A (en) | Method and device for filter processing | |
| EP2712176B1 (en) | Method and apparatus for photography | |
| CN104460185A (en) | Automatic focusing method and device | |
| CN106295530A (en) | Face identification method and device | |
| CN104933419A (en) | Method and device for obtaining iris images and iris identification equipment | |
| CN104867112B (en) | Photo processing method and device | |
| CN105335684A (en) | Face detection method and device | |
| CN108154466A (en) | Image processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20161116 |