CN102955947A - Equipment and method for determining image definition - Google Patents
Equipment and method for determining image definition Download PDFInfo
- Publication number
- CN102955947A CN102955947A CN2011102406285A CN201110240628A CN102955947A CN 102955947 A CN102955947 A CN 102955947A CN 2011102406285 A CN2011102406285 A CN 2011102406285A CN 201110240628 A CN201110240628 A CN 201110240628A CN 102955947 A CN102955947 A CN 102955947A
- Authority
- CN
- China
- Prior art keywords
- image
- sharpness
- equipment
- definition
- described image
- 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
Images
Abstract
The invention provides equipment and a method for determining image definition. The method includes: acquiring an image to be subjected to image definition determining; acquiring low-frequency noise corresponding to the image according to the image; and based on an SVM (support vector machine) definition classifier, determining the image definition of the image according to the low-frequency noise. Compared with the prior art, the equipment and the method have the advantages that when image query results are sequenced, presetting of any reference models is not needed, and by means of acquiring the image definition of the image query results, sharp images and blurred images in the image query results are resequenced based on the image definition, so that image search experience of users is improved.
Description
Technical field
The present invention relates to Internet technical field, relate in particular to for the image processing techniques of determining image definition.
Background technology
In the application such as search engine or image indexing system, image querying sequence is corresponding a plurality of image querying result probably, and the sequencing between these image queryings result often depends on various factors.In the prior art, for described image querying result's the order that provides is provided, must adopt predefined reference model that described image querying result is assessed in the time of most, then based on assessment result described image querying result be sorted.Yet, above-mentioned for assessment of the preset reference model data that need carry out numerous and complicated process and just can obtain, and different image querying the possibility of result can be suitable for different reference models, this also will cause reference model quantity huge, compatibility is lower.
In view of this, how to design a kind of method for determining image definition, need not to preset any reference model, by obtaining described image querying result's image definition, described image querying result resequences, the picture search that promotes the user is experienced, and is the problem that person skilled needs to be resolved hurrily.
Summary of the invention
The purpose of this invention is to provide a kind of Apparatus for () and method therefor for determining image definition.
According to an aspect of the present invention, provide a kind of computer implemented method for determining image definition, wherein, the method may further comprise the steps:
A obtains the image of image definition to be determined;
B obtains the low-frequency noise corresponding with described image according to described image;
C according to described low-frequency noise, determines the image definition of described image based on SVM sharpness sorter.
According to another aspect of the present invention, also provide a kind of equipment for determining image definition, wherein, described equipment comprises:
The first deriving means is for the image that obtains image definition to be determined;
The second deriving means is used for according to described image, obtains the low-frequency noise corresponding with described image;
First determines device, is used for based on SVM sharpness sorter, according to described low-frequency noise, determines the image definition of described image.
According to a further aspect of the invention, also provide a kind of search engine, wherein, this search engine comprises such as the described equipment for determining image definition of one aspect of the present invention.
Compared with prior art, the present invention is to the image querying sort result time, need not to preset any reference model, by obtaining described image querying result's image definition, and based on described image definition resequence picture rich in detail and blurred picture among the described image querying result, the picture search that promotes the user is experienced.
Description of drawings
By reading the detailed description that non-limiting example is done of doing with reference to the following drawings, it is more obvious that other features, objects and advantages of the present invention will become:
Fig. 1 illustrates the equipment synoptic diagram that is used for determining image definition according to one aspect of the invention;
Fig. 2 illustrates the equipment synoptic diagram that is used for determining image definition according to one embodiment of the present invention;
Fig. 3 illustrates the method flow diagram that is used for determining image definition according to another aspect of the present invention;
Fig. 4 illustrates the method flow diagram that is used for determining image definition according to one embodiment of the present invention.
Same or analogous Reference numeral represents same or analogous parts in the accompanying drawing.
Embodiment
Below in conjunction with accompanying drawing the present invention is described in further detail.
Fig. 1 illustrates the equipment synoptic diagram that is used for determining image definition according to one aspect of the invention.Wherein, determine that equipment 1 includes but not limited to the cloud that network host, single network server, a plurality of webserver collection or a plurality of server consist of.At this, cloud can be made of a large amount of computing machines or the webserver based on cloud computing (Cloud Computing), and wherein, cloud computing is a kind of super virtual machine that is comprised of the loosely-coupled computing machine collection of a group of Distributed Calculation.Wherein, described definite equipment 1 comprises the first deriving means 11, the second deriving means 12 and first definite device 13.
The first deriving means 11 obtains the image of image definition to be determined.Particularly, the first deriving means 11 obtains the image of described image definition to be determined such as the communication mode of the application programming interfaces (API) that provide by third party's equipment such as search engines or agreement from this third party device; Perhaps, the communication mode of the application programming interfaces (API) that provide by third party's equipment such as search engines or agreement, obtain this user by the image querying sequence of subscriber equipment input from this third party device, and described image querying sequence carried out matching inquiry in the search index storehouse, obtain the image of the described to be determined image definition corresponding with described image querying sequence; Perhaps, pass through page technology, such as ASP, JSP, PHP etc., obtain the user by the image querying sequence of this subscriber equipment input from subscriber equipment, and described image querying sequence carried out matching inquiry in the search index storehouse, obtain the image of the described to be determined image definition corresponding with described image querying sequence.For example, the user keys in search sequence " fresh flower " in the search input field, the application programming interfaces (API) that provide such as third party's equipment such as search engines or the communication mode of other agreements are provided the first deriving means 11, obtain described search sequence " fresh flower ", and obtain corresponding one or more images, such as rose picture, carnation image etc. according to the search sequence of obtaining " fresh flower ".And for example, for image indexing system, the first deriving means 11 is received from the image that newly is added into this image indexing system that this image indexing system sends, and perhaps chooses at random an image in different image category, and with the image of described image as image definition to be determined.Those skilled in the art will be understood that the above-mentioned mode of obtaining the image of image definition to be determined only is for example; the mode of other existing or images that obtain image definition to be determined that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
The second deriving means 12 obtains the low-frequency noise corresponding with described image according to described image.Particularly, the image that the second deriving means 12 obtains according to the first deriving means 11 is by obtaining low-frequency noise such as the pixel value difference of the noise spot that obtains described image and the quantity of described noise spot.For example, the second deriving means 12 can according to described image, be added up in the neighborhood of pixels, a certain pixel is respectively on x direction and y direction, with the difference of neighbor pixel, if this difference in the threshold range that predetermined noise threshold limits, then keeps described pixel value difference; If this difference exceeds the threshold range that described predetermined noise threshold limits, then give up described pixel value difference.For example, a certain pixel and the neighbor pixel pixel value difference on the x direction is 5, and the pixel value difference on the y direction is 8, and the described predetermined noise threshold scope on x direction and y direction is 2～10, then the pixel value difference on x direction and y direction remains with described pixel and described pixel, then add up respectively that all are retained average and the variance of the pixel value difference of pixel on x direction and y direction that gets off in this pixel field, with as the low-frequency noise corresponding with described image.At this, the pixel value difference on x direction and the y direction meets the requirements and is retained the pixel that gets off and is called noise spot, and the quantity of the pixel under the pixel value difference on x direction and the y direction meets the requirements and is retained is called the quantity of noise spot.And for example, the second deriving means 12 also according to the quantity of described pixel value difference and described noise spot, calculates the variance corresponding with described noise spot, and utilizes the variance of calculating to obtain the low-frequency noise corresponding with described image.Those skilled in the art will be understood that the above-mentioned mode of the low-frequency noise corresponding with described image of obtaining is only for giving an example; other existing or modes of obtaining the low-frequency noise corresponding with described image that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
First definite device 13 is determined the image definition of described image based on SVM sharpness sorter according to described low-frequency noise.Particularly, determine equipment 1 end or comprising SVM sharpness sorter 14 with network equipment end that described definite equipment 1 is connected by network, first determines that device 13 is with the input feature vector of described low-frequency noise as described SVM sharpness sorter 14, by such as inquiry mode or other matching ways, compare with the image definition in the described SVM sharpness sorter 14, thereby obtain the image definition of described image.At this, determine equipment 1 and comprise that the network between the network equipment end of described SVM sharpness sorter 14 includes but not limited to internet, wide area network, Metropolitan Area Network (MAN), LAN (Local Area Network), VPN network, wireless self-organization network (Ad Hoc network) etc.In addition, determine that the data transmission between equipment 1 and the described network equipment end includes but not limited to ICP/IP protocol, udp protocol etc.Those skilled in the art will be understood that the mode of the image definition of above-mentioned definite described image only is for example; the mode of the image definition of other definite described images existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, determine equipment 1 each the device between work continuously.Particularly, the first deriving means 11 obtains the image of image definition to be determined; The second deriving means 12 obtains the low-frequency noise corresponding with described image according to described image; First definite device 13 according to described low-frequency noise, is determined the image definition of described image based on SVM sharpness sorter.At this, it will be understood by those skilled in the art that each device that " continuing " refer to determine equipment 1 respectively according to the mode of operation of setting or adjust in real time require to carry out the obtaining of the obtaining of image of image definition to be determined, the low-frequency noise corresponding with described image, based on the determining of the described image definition of SVM sharpness sorter, until definite equipment 1 stops to obtain the image of image definition to be determined in a long time.
Preferably, described definite equipment 1 also comprises the trainer (not shown), according to the image of a large amount of demarcation sharpness, carries out the sharpness classification based training, to obtain described SVM sharpness sorter.Particularly, described first determines device 13 based on SVM sharpness sorter 14, and the low-frequency noise of obtaining according to the second deriving means 12 is determined the image definition of described image.At this, described SVM sharpness sorter 14 can obtain by the great amount of images of demarcating sharpness is carried out the sharpness classification based training, and needn't be associated between the image of the described great amount of images that participates in the sharpness classification based training and image definition to be determined.For example, described SVM sharpness sorter 14 comprises a plurality of input interfaces, each input interface receives the image feature information of image accordingly, and an input interface in described a plurality of input interface is used for receiving the low-frequency noise of image, when the described low-frequency noise of obtaining when described the second deriving means 12 inputed to described SVM sharpness sorter 14, described first determined that device 13 determines the image definition of described images based on described SVM sharpness sorter 14.Those skilled in the art will be understood that the above-mentioned mode of described SVM sharpness sorter of obtaining is only for giving an example; other existing or modes of obtaining described SVM sharpness sorter that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, described definite equipment 1 comprises that also second determines the device (not shown), and described second determines device according to described image definition, in conjunction with default clarity threshold, determines the clear grade corresponding with described image.Particularly, second definite device is determined device 13 determined image definitions according to described first, described image definition and default clarity threshold is compared, to determine the clear grade of described image.For example, when described clarity threshold is single threshold value, the image of determining image definition can be divided into picture rich in detail and blurred picture two-stage, namely, when described image definition during more than or equal to described clarity threshold, the image corresponding with described image definition is picture rich in detail, and when described image definition during less than described clarity threshold, the image corresponding with described image definition is blurred picture.And for example, when described clarity threshold is dual thresholds (such as 0.5 and 0.8), the image of determining image definition can be divided into picture rich in detail, than three grades of picture rich in detail and blurred pictures, namely, when described image definition more than or equal to 0.8 the time, the image corresponding with described image definition is picture rich in detail; When described image definition less than 0.5 the time, the image corresponding with described image definition is blurred picture; And when described image definition was between 0.5 and 0.8, the image corresponding with described image definition was than picture rich in detail.From the above, the present invention determines that with first device 13 determined image definitions are in conjunction with default clarity threshold, can be some grades with the image subdivision corresponding with described image definition, when the user uses the image querying sequence to search for, can freely select to be positioned among the image querying result corresponding with described image querying sequence all images of a certain clear grade, strengthen user's search experience, also promoted the intelligent search ability of search engine.Those skilled in the art will be understood that and above-mentionedly determine that the mode of the clear grade corresponding with described image is only for giving an example; other existing or modes of determining the clear grade corresponding with described image that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, described definite equipment 1 also comprises the 4th deriving means (not shown) and the 3rd definite device (not shown), described the 4th deriving means carries out Region Segmentation to described image, to obtain a plurality of regional subimage corresponding with described image; With at least one image as described sharpness to be determined in described a plurality of regional subimages, utilize described the second deriving means 12 and described first to determine device 13, obtain in described a plurality of regional subimage the regional sharpness of at least one; Wherein, the 3rd determines device according to described regional sharpness, and the image definition of described image is determined in weighting.Particularly, described the 4th deriving means is such as carrying out Region Segmentation take 4 * 4 (pixels) or 16 * 16 (pixels) as unit with the image of image definition to be determined, thereby obtains a plurality of regional subimage corresponding with described image; Then with the image of each regional subimage as sharpness to be determined, obtain the low-frequency noise corresponding with described regional subimage by the second deriving means 12, and the regional sharpness of determining described regional subimage by first definite device 13 based on described low-frequency noise.The described the 3rd determines device according to the determined regional sharpness of described the 4th deriving means, and the image definition of described image is determined in weighting.For example, the described the 3rd determines that device comes weighting to determine the image definition of described image with following formula:
S＝E
c*weight
c+E
s*weight
s+E
l*weight
l
Wherein, S represents the image definition of described image, weight
cThe corresponding regional sharpness E in presentation video central area
cWeight, weight
sThe corresponding regional sharpness E of four fringe regions of presentation video
sWeight, weight
lThe corresponding regional sharpness E of four corner areas of presentation video
lWeight.People are when watching image, and the concerned degree of this image regional is different, and usually, the attention rate in picture centre zone is the highest, and the attention rate of four fringe regions of image is taken second place, and the attention rate of four corner areas of image is minimum, that is, and and weight
c＞weight
s＞weight
l
In a preferred embodiment (with reference to Fig. 1), described definite equipment 1 also comprises the generator (not shown), when described the first deriving means 11 obtain with the user by the corresponding image of the image querying sequence of subscriber equipment input after, described generator offers described subscriber equipment with described image and the image definition corresponding with described image.Referring to Fig. 1 the preferred embodiment is described in detail, wherein, the first deriving means 11 obtains the image by the corresponding image definition to be determined of the image querying sequence of subscriber equipment input with the user, the second deriving means 12 obtains the low-frequency noise corresponding with described image according to described image, first determines that device 13 is based on SVM sharpness sorter, determine the image definition of described image according to described low-frequency noise, generator offers described subscriber equipment with described image and the image definition corresponding with described image.Its detailed process determines that with reference to the second deriving means 12 among the described embodiment of Fig. 1 and first the performed process of device 13 is identical with aforementioned, for simplicity's sake, is contained in this with way of reference, does not give unnecessary details and do not do.Particularly, the first deriving means 11 is by the application programming interfaces (API) that provide such as third party's equipment such as search engines or the communication mode of other agreements are provided, obtain image corresponding to user images search sequence, and described generator determines that with described image and by described first device 13 determined image definitions offer described subscriber equipment.At this, the user carries out mutual mode by described subscriber equipment and definite equipment 1, include but not limited to keyboard, mouse, telepilot, touch pad, handwriting equipment or voice-input device, input picture search sequence in the input frame of browser software, application program or client software etc.
Preferably, described definite equipment 1 also comprises the collator (not shown), and described collator is according to described image definition, and the text-dependent degree in conjunction with corresponding with described image sorts to described image; Wherein, described generator also offers described subscriber equipment with described image after sorted.Particularly, described collator is determined device 13 determined image definitions and described text-dependent degree according to described first, described image is sorted, and described generator offers described subscriber equipment with image after sorted.At this, described text-dependent degree is the text relevant between described image querying sequence and the described image.The image definition of described image is combined with described text-dependent degree, when the user searches in the input picture search sequence, always preferentially see image definition higher and with the maximally related image of described image querying sequence, thereby the picture search that makes search engine can also promote the user when guaranteeing search accuracy rate is experienced.
More preferably, described definite equipment 1 also comprises the 5th deriving means (not shown), and described the 5th deriving means obtains the sharpness demand information corresponding with described image querying sequence according to the image definition statistical study of described image; Wherein, described collator, sorts to described image in conjunction with described text-dependent degree and described sharpness demand information also according to described image definition.For example, different image querying sequences can be corresponding to different image definition requirements, described the 5th deriving means can be according to the image definition statistical study of image, obtain the sharpness demand information corresponding with described image querying sequence, then described collator sorts to described image based on described image definition, described text-dependent degree and described sharpness demand information.For example, when user input query sequence " fresh flower ", image definition to the image corresponding with described search sequence is carried out statistical study, if the image definition of described image is all higher, search engine just need to sort with this dimension of image definition when image is provided to the user.
Fig. 2 illustrates the equipment synoptic diagram that is used for determining image definition according to one preferred embodiment of the present invention.Wherein, determine that equipment 1 ' includes but not limited to the cloud that network host, single network server, a plurality of webserver collection or a plurality of server consist of.At this, cloud can be made of a large amount of computing machines or the webserver based on cloud computing (Cloud Computing), and wherein, cloud computing is a kind of super virtual machine that is comprised of the loosely-coupled computing machine collection of a group of Distributed Calculation.Wherein, described definite equipment 1 ' comprises the first deriving means 11 ', the second deriving means 12 ', the first definite device 13 ' and the 3rd deriving means 15 '.
In really locking equipment 1 ' shown in Figure 2, the first deriving means 11 ', the second deriving means 12 ', first determine that device 13 ' is same or similar with the first deriving means 11, the second deriving means 12, first definite device 13 shown in Figure 1 respectively, for describing for simplicity, so locate to repeat no more, and mode by reference is contained in this.
The 3rd deriving means 15 ' obtains the image auxiliary information corresponding with described image according to described image; And described first definite device 13 ' according to described low-frequency noise and described image auxiliary information, is determined the image definition of described image also based on described SVM sharpness sorter.Particularly, the 3rd deriving means 15 ' by such as filtering, convergent-divergent or other modes described image being processed, obtains the image auxiliary information corresponding with described image according to described image.First determines that image auxiliary information that low-frequency noise that device 13 ' obtains the second deriving means 12 ' and the 3rd deriving means 15 ' obtain is as the input feature vector of two different dimensions in the SVM sharpness sorter, by such as inquiry mode or other matching ways, compare with the image definition in the described SVM sharpness sorter 14 ', thereby obtain the image definition of described image.Those skilled in the art will be understood that the above-mentioned mode of the image auxiliary information corresponding with described image of obtaining is only for giving an example; other existing or modes of obtaining the image auxiliary information corresponding with described image that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, described image auxiliary information comprise following at least each:
The fuzzy difference of-described image;
-described imagezoom difference;
The marginal sharpness of-described image.
Comprise that take described image auxiliary information the fuzzy difference of described image is as example, described the 3rd deriving means 15 ' carries out filtering with the image of the image definition to be determined that the first deriving means 11 ' obtains, to obtain the blurred picture after filtering is processed, described blurred picture and each self-corresponding pixel of described image are carried out subtraction, obtain pixel value difference, obtain the fuzzy difference corresponding with described image thereby the absolute value phase adduction of these pixel value differences is averaging.First definite device 13 ' according to described low-frequency noise and described fuzzy difference, is determined the image definition of described image based on described SVM sharpness sorter.At this, described image is carried out filtering include but not limited to: gaussian filtering, medium filtering or the combination of the two.Comprise that take described image auxiliary information described imagezoom difference is as example, described the 3rd deriving means 15 ' carries out convergent-divergent with the image of the image definition to be determined that the first deriving means 11 ' obtains, to obtain the zoomed image after convergent-divergent is processed, and described zoomed image and each self-corresponding pixel of described image carried out subtraction, obtain pixel value difference, obtain the convergent-divergent difference corresponding with described image thereby the absolute value phase adduction of these pixel value differences is averaging.At this, described image is carried out convergent-divergent include but not limited to: neighborhood sampling method, cube convolution method or the combination of the two.First definite device 13 ' according to described low-frequency noise and described convergent-divergent difference, is determined the image definition of described image based on described SVM sharpness sorter.The marginal sharpness that comprises described image take described image auxiliary information is example, and in general, grey scale change Shaoxing opera is strong, and the image border is more clear, and the sharpness of image is also just higher.For example, the 3rd deriving means 15 ' can be added up the grey scale change of described image border normal direction according to described image, then obtains the marginal sharpness of described image based on described grey scale change.First definite device 13 ' according to described low-frequency noise and described marginal sharpness, is determined the image definition of described image based on described SVM sharpness sorter.
Preferably, described definite equipment 1 ' also comprises the 4th deriving means (not shown), and described the 4th deriving means carries out Region Segmentation to described image, to obtain a plurality of regional subimage corresponding with described image; With at least one image as described sharpness to be determined in described a plurality of regional subimages, utilize described the second deriving means 12 ' and described first to determine device 13 ', obtain in described a plurality of regional subimage the regional sharpness of at least one; Wherein, described definite equipment 1 ' comprises that also the 3rd determines the device (not shown), and the described the 3rd determines device according to described regional sharpness, and the image definition of described image is determined in weighting.Particularly, described the 4th deriving means is such as carrying out Region Segmentation take 4 * 4 (pixels) or 16 * 16 (pixels) as unit with the image of image definition to be determined, thereby obtains a plurality of regional subimage corresponding with described image; Then with the image of each regional subimage as sharpness to be determined, obtain the low-frequency noise corresponding with described regional subimage by the second deriving means 12 ', and the regional sharpness of determining described regional subimage by first definite device 13 ' based on described low-frequency noise.The described the 3rd determines device according to the determined regional sharpness of described the 4th deriving means, and the image definition of described image is determined in weighting.For example, the described the 3rd determines that device comes weighting to determine the image definition of described image with following formula:
S＝E
c*weight
c+E
s*weight
s+E
l*weight
l
Wherein, S represents the image definition of described image, weight
cThe corresponding regional sharpness E in presentation video central area
cWeight, weight
sThe corresponding regional sharpness E of four fringe regions of presentation video
sWeight, weight
lThe corresponding regional sharpness E of four corner areas of presentation video
lWeight.People are when watching image, and the concerned degree of this image regional is different, and usually, the attention rate in picture centre zone is the highest, and the attention rate of four fringe regions of image is taken second place, and the attention rate of four corner areas of image is minimum, that is, and and weight
c＞weight
s＞weight
l
In addition, above-mentioned equipment for determining image definition can combine with existing search engine, consists of a kind of new search engine, and existing search engine can adopt known to search engines such as Baidu, Google, Yahoo.
Preferably, described search engine will offer described subscriber equipment by the corresponding image of the image querying sequence of subscriber equipment input and with the corresponding image definition of described image with the user.Particularly, described search engine also by modes such as special font, floating frame, provides the image definition corresponding with described image when providing with the user by the corresponding image of the image querying sequence of subscriber equipment input.For example, combine with existing search engine, when the image querying sequence of inputting by subscriber equipment according to the user provides corresponding image, in the explanation of these images, add corresponding image definition; Further, the image definition of these images especially mode such as font or floating frame shows, when resting on image or picture specification such as the mouse the user, shows the image definition of this image with the suspension window.Those skilled in the art will be understood that the presentation mode of above-mentioned image definition is only for giving an example; the presentation mode of other image definitions existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Fig. 3 illustrates the method flow diagram that is used for determining image definition according to a further aspect of the present invention.Wherein, be used for to determine image definition really locking equipment include but not limited to the cloud that network host, single network server, a plurality of webserver collection or a plurality of server consist of.At this, cloud can be made of a large amount of computing machines or the webserver based on cloud computing (Cloud Computing), and wherein, cloud computing is a kind of super virtual machine that is comprised of the loosely-coupled computing machine collection of a group of Distributed Calculation.
In step S1, described definite equipment obtains the image of image definition to be determined.Particularly, described definite equipment obtains the image of described image definition to be determined such as the communication mode of the application programming interfaces (API) that provide by third party's equipment such as search engines or agreement from this third party device; Perhaps, the communication mode of the application programming interfaces (API) that provide by third party's equipment such as search engines or agreement, obtain this user by the image querying sequence of subscriber equipment input from this third party device, and described image querying sequence carried out matching inquiry in the search index storehouse, obtain the image of the described to be determined image definition corresponding with described image querying sequence; Perhaps, pass through page technology, such as ASP, JSP, PHP etc., obtain the user by the image querying sequence of this subscriber equipment input from subscriber equipment, and described image querying sequence carried out matching inquiry in the search index storehouse, obtain the image of the described to be determined image definition corresponding with described image querying sequence.For example, the user keys in search sequence " fresh flower " in the search input field, the application programming interfaces (API) that described definite equipment calls such as third party's equipment such as search engine provide or the communication mode of other agreements, obtain described search sequence " fresh flower ", and obtain corresponding one or more images, such as rose picture, carnation image etc. according to the search sequence of obtaining " fresh flower ".And for example, for image indexing system, described definite equipment is received from the image that newly is added into this image indexing system that this image indexing system sends, and perhaps chooses at random an image in different image category, and with the image of described image as image definition to be determined.Those skilled in the art will be understood that the above-mentioned mode of obtaining the image of image definition to be determined only is for example; the mode of other existing or images that obtain image definition to be determined that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
In step S2, described definite equipment obtains the low-frequency noise corresponding with described image according to described image.Particularly, the image that described definite equipment obtains according to above-mentioned steps S1 is by obtaining low-frequency noise such as the pixel value difference of the noise spot that obtains described image and the quantity of described noise spot.For example, described definite equipment can according to described image, be added up in the neighborhood of pixels, a certain pixel is respectively on x direction and y direction, with the difference of neighbor pixel, if this difference in the threshold range that predetermined noise threshold limits, then keeps described pixel value difference; If this difference exceeds the threshold range that described predetermined noise threshold limits, then give up described pixel value difference.For example, a certain pixel and the neighbor pixel pixel value difference on the x direction is 5, and the pixel value difference on the y direction is 8, and the described predetermined noise threshold scope on x direction and y direction is 2～10, then the pixel value difference on x direction and y direction remains with described pixel and described pixel, then add up respectively that all are retained average and the variance of the pixel value difference of pixel on x direction and y direction that gets off in this pixel field, with as the low-frequency noise corresponding with described image.At this, the pixel value difference on x direction and the y direction meets the requirements and is retained the pixel that gets off and is called noise spot, and the quantity of the pixel under the pixel value difference on x direction and the y direction meets the requirements and is retained is called the quantity of noise spot.And for example, described definite equipment also according to the quantity of described pixel value difference and described noise spot, calculates the variance corresponding with described noise spot, and utilizes the variance of calculating to obtain the low-frequency noise corresponding with described image.Those skilled in the art will be understood that the above-mentioned mode of the low-frequency noise corresponding with described image of obtaining is only for giving an example; other existing or modes of obtaining the low-frequency noise corresponding with described image that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
In step S3, described definite equipment is determined the image definition of described image based on SVM sharpness sorter according to described low-frequency noise.Particularly, comprise SVM sharpness sorter in described definite equipment end or with network equipment end that described definite equipment is connected by network, described definite equipment is with the input feature vector of described low-frequency noise as described SVM sharpness sorter, by such as inquiry mode or other matching ways, compare with the image definition in the described SVM sharpness sorter, thereby obtain the image definition of described image.At this, described definite equipment and comprise that the network between the network equipment end of described SVM sharpness sorter includes but not limited to internet, wide area network, Metropolitan Area Network (MAN), LAN (Local Area Network), VPN network, wireless self-organization network (Ad Hoc network) etc.In addition, the data between described definite equipment and the described network equipment end transmit and include but not limited to ICP/IP protocol, udp protocol etc.Those skilled in the art will be understood that the mode of the image definition of above-mentioned definite described image only is for example; the mode of the image definition of other definite described images existing or that may occur from now on is as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, the method also comprises step S5 (not shown), and in described step S5, described definite equipment carries out the sharpness classification based training, to obtain described SVM sharpness sorter according to a large amount of images of demarcating sharpness.Particularly, described definite equipment is determined the image definition of described image based on SVM sharpness sorter 14 according to the low-frequency noise of obtaining.At this, described SVM sharpness sorter can obtain by the great amount of images of demarcating sharpness is carried out the sharpness classification based training, and needn't be associated between the image of the described great amount of images that participates in the sharpness classification based training and image definition to be determined.For example, described SVM sharpness sorter comprises a plurality of input interfaces, each input interface receives the image feature information of image accordingly, and an input interface in described a plurality of input interface is used for receiving the low-frequency noise of image, when the low-frequency noise of obtaining when described definite equipment inputed to described SVM sharpness sorter, described definite equipment was determined the image definition of described image based on described SVM sharpness sorter.Those skilled in the art will be understood that the above-mentioned mode of described SVM sharpness sorter of obtaining is only for giving an example; other existing or modes of obtaining described SVM sharpness sorter that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, the method also comprises step S6 (not shown), and in described step S6, described definite equipment in conjunction with default clarity threshold, is determined the clear grade corresponding with described image according to described image definition.Particularly, described definite equipment is compared described image definition and default clarity threshold according to the determined image definition of described step S3, to determine the clear grade of described image.For example, when described clarity threshold is single threshold value, the image of determining image definition can be divided into picture rich in detail and blurred picture two-stage, namely, when described image definition during more than or equal to described clarity threshold, the image corresponding with described image definition is picture rich in detail, and when described image definition during less than described clarity threshold, the image corresponding with described image definition is blurred picture.And for example, when described clarity threshold is dual thresholds (such as 0.5 and 0.8), the image of determining image definition can be divided into picture rich in detail, than three grades of picture rich in detail and blurred pictures, namely, when described image definition more than or equal to 0.8 the time, the image corresponding with described image definition is picture rich in detail; When described image definition less than 0.5 the time, the image corresponding with described image definition is blurred picture; And when described image definition was between 0.5 and 0.8, the image corresponding with described image definition was than picture rich in detail.From the above, the present invention is with the default clarity threshold of the determined image definition combination of described definite equipment, can be some grades with the image subdivision corresponding with described image definition, when the user uses the image querying sequence to search for, can freely select to be positioned among the image querying result corresponding with described image querying sequence all images of a certain clear grade, strengthen user's search experience, also promoted the intelligent search ability of search engine.Those skilled in the art will be understood that and above-mentionedly determine that the mode of the clear grade corresponding with described image is only for giving an example; other existing or modes of determining the clear grade corresponding with described image that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, the method also comprises step S7 (not shown), and in described step S7, described definite equipment carries out Region Segmentation to described image, to obtain a plurality of regional subimage corresponding with described image; And with at least one image as described sharpness to be determined in described a plurality of regional subimages, utilize described step S2 and described step S3, obtain in described a plurality of regional subimage the regional sharpness of at least one; Wherein, described definite equipment is according to described regional sharpness, and the image definition of described image is determined in weighting.Particularly, described definite equipment is such as carrying out Region Segmentation take 4 * 4 (pixels) or 16 * 16 (pixels) as unit with the image of image definition to be determined, thereby obtains a plurality of regional subimage corresponding with described image; Then with the image of each regional subimage as sharpness to be determined, obtain the low-frequency noise corresponding with described regional subimage by step S2, and the regional sharpness of determining described regional subimage by step S3 based on described low-frequency noise.Described definite equipment is according to determined regional sharpness, and the image definition of described image is determined in weighting.For example, described definite equipment comes weighting to determine the image definition of described image with following formula:
S＝E
c*weight
c+E
s*weight
s+E
l*weight
l
Wherein, S represents the image definition of described image, weight
cThe corresponding regional sharpness E in presentation video central area
cWeight, weight
sThe corresponding regional sharpness E of four fringe regions of presentation video
sWeight, weight
lThe corresponding regional sharpness E of four corner areas of presentation video
lWeight.People are when watching image, and the concerned degree of this image regional is different, and usually, the attention rate in picture centre zone is the highest, and the attention rate of four fringe regions of image is taken second place, and the attention rate of four corner areas of image is minimum, that is, and and weight
c＞weight
s＞weight
l
In a preferred embodiment (with reference to Fig. 3), the method also comprises step S8 (not shown), in described step S8, described definite equipment will offer described subscriber equipment by the corresponding described image of the image querying sequence of subscriber equipment input and with the corresponding image definition of described image with the user.Referring to Fig. 3 the preferred embodiment is described in detail, wherein, in step S1, described definite equipment obtains the image by the corresponding image definition to be determined of the image querying sequence of subscriber equipment input with the user, in step S2, described definite equipment obtains the low-frequency noise corresponding with described image according to described image, in step S3, described definite equipment is based on SVM sharpness sorter, determine the image definition of described image according to described low-frequency noise, in described step S8, described definite equipment will offer described subscriber equipment by the corresponding described image of the image querying sequence of subscriber equipment input and with the corresponding image definition of described image with the user.Its detailed process for simplicity's sake, is contained in this with way of reference with aforementioned identical with reference to the performed process of step S2 described in the described embodiment of Fig. 3 and described step S3, does not give unnecessary details and do not do.Particularly, described definite equipment is by the application programming interfaces (API) that provide such as third party's equipment such as search engines or the communication mode of other agreements are provided, obtain image corresponding to user images search sequence, and described image and the image definition corresponding with described image are offered described subscriber equipment.At this, the user carries out mutual mode by described subscriber equipment and definite equipment, include but not limited to keyboard, mouse, telepilot, touch pad, handwriting equipment or voice-input device, input picture search sequence in the input frame of browser software, application program or client software etc.
Preferably, the method also comprises step S9 (not shown), and in described step S9, described definite equipment is according to described image definition, and the text-dependent degree in conjunction with corresponding with described image sorts to described image; Then described image is after sorted offered described subscriber equipment.Particularly, described definite equipment sorts to described image according to the determined image definition of described step S3 and described text-dependent degree, and image is after sorted offered described subscriber equipment.At this, described text-dependent degree is the text relevant between described image querying sequence and the described image.The image definition of described image is combined with described text-dependent degree, when the user searches in the input picture search sequence, always preferentially see image definition higher and with the maximally related image of described image querying sequence, thereby the picture search that makes search engine can also promote the user when guaranteeing search accuracy rate is experienced.
More preferably, the method also comprises step S10 (not shown), and in described step S10, described definite equipment obtains the sharpness demand information corresponding with described image querying sequence according to the image definition statistical study of described image; Then according to described image definition, in conjunction with described text-dependent degree and described sharpness demand information, described image is sorted.For example, different image querying sequences can be corresponding to different image definition requirements, described definite equipment can be according to the image definition statistical study of image, obtain the sharpness demand information corresponding with described image querying sequence, then based on described image definition, described text-dependent degree and described sharpness demand information, described image is sorted.For example, when user input query sequence " fresh flower ", image definition to the image corresponding with described search sequence is carried out statistical study, if the image definition of described image is all higher, search engine just need to sort with this dimension of image definition when image is provided to the user.
Fig. 4 illustrates the method flow diagram that is used for determining image definition according to one preferred embodiment of the present invention.Wherein, be used for to determine image definition really locking equipment include but not limited to the cloud that network host, single network server, a plurality of webserver collection or a plurality of server consist of.At this, cloud can be made of a large amount of computing machines or the webserver based on cloud computing (Cloud Computing), and wherein, cloud computing is a kind of super virtual machine that is comprised of the loosely-coupled computing machine collection of a group of Distributed Calculation.
With reference to Fig. 4, step S1 ', step S2 ' and step S3 ' are same or similar with step S1, step S2, step S3 shown in Figure 3 respectively, for describing for simplicity, so locate to repeat no more, and mode by reference is contained in this.
In step S5 ', described definite equipment obtains the image auxiliary information corresponding with described image according to described image; And based on described SVM sharpness sorter, according to described low-frequency noise and described image auxiliary information, determine the image definition of described image.Particularly, described definite equipment by such as filtering, convergent-divergent or other modes described image being processed, obtains the image auxiliary information corresponding with described image according to described image.Described definite equipment will utilize image auxiliary information that low-frequency noise that described step S2 ' obtains and described step S5 ' obtain as the input feature vector of two different dimensions in the SVM sharpness sorter, by such as inquiry mode or other matching ways, compare with the image definition in the described SVM sharpness sorter, thereby obtain the image definition of described image.Those skilled in the art will be understood that the above-mentioned mode of the image auxiliary information corresponding with described image of obtaining is only for giving an example; other existing or modes of obtaining the image auxiliary information corresponding with described image that may occur from now on are as applicable to the present invention; also should be included in the protection domain of the present invention, and be contained in this with way of reference.
Preferably, described image auxiliary information comprise following at least each:
The fuzzy difference of-described image;
-described imagezoom difference;
The marginal sharpness of-described image.
Comprise that take described image auxiliary information the fuzzy difference of described image is as example, in step S5 ', described definite equipment will utilize the image of the image definition to be determined that described step S1 ' obtains to carry out filtering, to obtain the blurred picture after filtering is processed, described blurred picture and each self-corresponding pixel of described image are carried out subtraction, obtain pixel value difference, obtain the fuzzy difference corresponding with described image thereby the absolute value phase adduction of these pixel value differences is averaging.Described definite equipment according to described low-frequency noise and described fuzzy difference, is determined the image definition of described image based on described SVM sharpness sorter.At this, described image is carried out filtering include but not limited to: gaussian filtering, medium filtering or the combination of the two.Comprise that take described image auxiliary information described imagezoom difference is as example, in step S5 ', described definite equipment will utilize the image of the image definition to be determined that described step S1 ' obtains to carry out convergent-divergent, to obtain the zoomed image after convergent-divergent is processed, and described zoomed image and each self-corresponding pixel of described image carried out subtraction, obtain pixel value difference, obtain the convergent-divergent difference corresponding with described image thereby the absolute value phase adduction of these pixel value differences is averaging.At this, described image is carried out convergent-divergent include but not limited to: neighborhood sampling method, cube convolution method or the combination of the two.Described definite equipment according to described low-frequency noise and described convergent-divergent difference, is determined the image definition of described image based on described SVM sharpness sorter.The marginal sharpness that comprises described image take described image auxiliary information is example, and in general, grey scale change Shaoxing opera is strong, and the image border is more clear, and the sharpness of image is also just higher.For example, in step S5 ', described definite equipment is added up the grey scale change of described image border normal direction according to described image, then obtains the marginal sharpness of described image based on described grey scale change.Described definite equipment according to described low-frequency noise and described marginal sharpness, is determined the image definition of described image based on described SVM sharpness sorter.
Preferably, the method also comprises step S6 ' and step S7 ' (all not shown), and at described step S6 ', described definite equipment carries out Region Segmentation to described image, to obtain a plurality of regional subimage corresponding with described image; Then with at least one image as described sharpness to be determined in described a plurality of regional subimages, utilize described step S2 ' and described step S3 ', obtain in described a plurality of regional subimage the regional sharpness of at least one.In described step S7 ', described definite equipment is according to described regional sharpness, and the image definition of described image is determined in weighting.Particularly, described definite equipment is such as carrying out Region Segmentation take 4 * 4 (pixels) or 16 * 16 (pixels) as unit with the image of image definition to be determined, thereby obtains a plurality of regional subimage corresponding with described image; Then with the image of each regional subimage as sharpness to be determined, obtain the low-frequency noise corresponding with described regional subimage by step S2 ', and the regional sharpness of determining described regional subimage by step S3 ' based on described low-frequency noise.Described definite equipment is according to determined regional sharpness, and the image definition of described image is determined in weighting.For example, described definite equipment comes weighting to determine the image definition of described image with following formula:
S＝E
c*weight
c+E
s*weight
s+E
l*weight
l
Wherein, S represents the image definition of described image, weight
cThe corresponding regional sharpness E in presentation video central area
cWeight, weight
sThe corresponding regional sharpness E of four fringe regions of presentation video
sWeight, weight
lThe corresponding regional sharpness E of four corner areas of presentation video
lWeight.People are when watching image, and the concerned degree of this image regional is different, and usually, the attention rate in picture centre zone is the highest, and the attention rate of four fringe regions of image is taken second place, and the attention rate of four corner areas of image is minimum, that is, and and weight
c＞weight
s＞weight
l
To those skilled in the art, obviously the invention is not restricted to the details of above-mentioned example embodiment, and in the situation that does not deviate from spirit of the present invention or essential characteristic, can realize the present invention with other concrete form.Therefore, no matter from which point, all should regard embodiment as exemplary, and be nonrestrictive, scope of the present invention is limited by claims rather than above-mentioned explanation, therefore is intended to include in the present invention dropping on the implication that is equal to important document of claim and all changes in the scope.Any Reference numeral in the claim should be considered as limit related claim.In addition, obviously other unit or step do not got rid of in " comprising " word, and odd number is not got rid of plural number.A plurality of unit of stating in system's claim or device also can be realized by software or hardware by a unit or device.The first, the second word such as grade is used for representing title, and does not represent any specific order.
Claims (22)
1. computer implemented method for determining image definition, wherein, the method may further comprise the steps:
A obtains the image of image definition to be determined;
B obtains the low-frequency noise corresponding with described image according to described image;
C according to described low-frequency noise, determines the image definition of described image based on SVM sharpness sorter.
2. method according to claim 1, wherein, described method also comprises:
-according to a large amount of images of demarcating sharpness, carry out the sharpness classification based training, to obtain described SVM sharpness sorter.
3. method according to claim 1 and 2, wherein, described step b also comprises:
-according to predetermined noise threshold, determine the quantity of pixel value difference and the described noise spot of described noise in image point;
-according to the quantity of described pixel value difference and described noise spot, calculate the variance corresponding with described noise spot, to obtain described low-frequency noise.
4. each described method in 3 according to claim 1, wherein, described method also comprises:
-according to described image, obtain the image auxiliary information corresponding with described image;
Wherein, described step c also comprises:
-based on described SVM sharpness sorter, according to described low-frequency noise and described image auxiliary information, determine the image definition of described image.
5. method according to claim 4, wherein, described image auxiliary information comprise following at least each:
The fuzzy difference of-described image;
-described imagezoom difference;
The marginal sharpness of-described image.
6. each described method in 5 according to claim 1, wherein, described method also comprises:
-according to described image definition, in conjunction with default clarity threshold, determine the clear grade corresponding with described image.
7. each described method in 6 according to claim 1, wherein, described method also comprises step v:
-described image is carried out Region Segmentation, to obtain a plurality of regional subimage corresponding with described image;
-with at least one image as described sharpness to be determined in described a plurality of regional subimages, repeating step b and c are to obtain in described a plurality of regional subimage the regional sharpness of at least one;
Wherein, the method also comprises:
-according to described regional sharpness, the image definition of described image is determined in weighting.
8. each described method in 7 according to claim 1, wherein, described step a also comprises:
-obtain with the user by the corresponding image of the image querying sequence of subscriber equipment input;
Wherein, described method also comprises:
X offers described subscriber equipment with described image and the image definition corresponding with described image.
9. method according to claim 8, wherein, described method also comprises:
R is according to described image definition, and the text-dependent degree in conjunction with corresponding with described image sorts to described image;
Wherein, described step x also comprises:
-described image is after sorted offered described subscriber equipment.
10. method according to claim 9, wherein, described method also comprises:
-according to the image definition statistical study of described image, obtain the sharpness demand information corresponding with described image querying sequence;
Wherein, described step r also comprises:
-according to described image definition, in conjunction with described text-dependent degree and described sharpness demand information, described image is sorted.
11. an equipment that is used for determining image definition, wherein, described equipment comprises:
The first deriving means is for the image that obtains image definition to be determined;
The second deriving means is used for according to described image, obtains the low-frequency noise corresponding with described image;
First determines device, is used for based on SVM sharpness sorter, according to described low-frequency noise, determines the image definition of described image.
12. equipment according to claim 11, wherein, described equipment also comprises trainer, is used for: according to the image of a large amount of demarcation sharpness, carry out the sharpness classification based training, to obtain described SVM sharpness sorter.
13. according to claim 11 or 12 described equipment, wherein, described the second deriving means also is used for:
-according to predetermined noise threshold, determine the quantity of pixel value difference and the described noise spot of described noise in image point;
-according to the quantity of described pixel value difference and described noise spot, calculate the variance corresponding with described noise spot, to obtain described low-frequency noise.
14. each described equipment in 13 according to claim 11, wherein, described equipment also comprises the 3rd deriving means, is used for:
-according to described image, obtain the image auxiliary information corresponding with described image;
Wherein, described first determines that device also is used for:
-based on described SVM sharpness sorter, according to described low-frequency noise and described image auxiliary information, determine the image definition of described image.
15. equipment according to claim 14, wherein, described image auxiliary information comprise following at least each:
The fuzzy difference of-described image;
-described imagezoom difference;
The marginal sharpness of-described image.
16. each described equipment in 15 according to claim 11, wherein, described equipment comprises that also second determines device, is used for:
-according to described image definition, in conjunction with default clarity threshold, determine the clear grade corresponding with described image.
17. each described equipment in 16 according to claim 11, wherein, described equipment also comprises the 4th deriving means, is used for:
-described image is carried out Region Segmentation, to obtain a plurality of regional subimage corresponding with described image;
-with at least one image as described sharpness to be determined in described a plurality of regional subimages, utilize described the second deriving means and described first to determine device, obtain in described a plurality of regional subimage the regional sharpness of at least one;
Wherein, described equipment comprises that also the 3rd determines device, is used for:
-according to described regional sharpness, the image definition of described image is determined in weighting.
18. each described equipment in 17 according to claim 11, wherein, described the first deriving means also is used for:
-obtain with the user by the corresponding image of the image querying sequence of subscriber equipment input;
Wherein, described equipment also comprises generator, is used for:
-described image and the image definition corresponding with described image are offered described subscriber equipment.
19. equipment according to claim 18, wherein, described equipment also comprises collator, is used for according to described image definition, and the text-dependent degree in conjunction with corresponding with described image sorts to described image;
Wherein, described generator also is used for:
-described image is after sorted offered described subscriber equipment.
20. equipment according to claim 19, wherein, described equipment also comprises the 5th deriving means, is used for the image definition statistical study according to described image, obtains the sharpness demand information corresponding with described image querying sequence;
Wherein, described collator also is used for:
-according to described image definition, in conjunction with described text-dependent degree and described sharpness demand information, described image is sorted.
21. a search engine, wherein, this search engine comprises such as each described equipment for determining image definition in the claim 11 to 20.
22. search engine according to claim 21, wherein, this search engine also provides the image definition corresponding with described image when providing with the user by the corresponding image of the image querying sequence of subscriber equipment input.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110240628.5A CN102955947B (en) | 2011-08-19 | 2011-08-19 | A kind of device and method thereof for being used to determine image definition |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110240628.5A CN102955947B (en) | 2011-08-19 | 2011-08-19 | A kind of device and method thereof for being used to determine image definition |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102955947A true CN102955947A (en) | 2013-03-06 |
| CN102955947B CN102955947B (en) | 2017-09-22 |
Family
ID=47764732
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110240628.5A Active CN102955947B (en) | 2011-08-19 | 2011-08-19 | A kind of device and method thereof for being used to determine image definition |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102955947B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106713864A (en) * | 2016-12-24 | 2017-05-24 | 深圳云天励飞技术有限公司 | Video processing method and device |
| CN109801288A (en) * | 2019-01-25 | 2019-05-24 | 淮阴师范学院 | A kind of image Focus field emission array implementation method based on directional statistics characteristic |
| WO2019137167A1 (en) * | 2018-01-10 | 2019-07-18 | Oppo广东移动通信有限公司 | Photo album management method and apparatus, storage medium, and electronic device |
| CN110175980A (en) * | 2019-04-11 | 2019-08-27 | 平安科技（深圳）有限公司 | Image definition recognition methods, image definition identification device and terminal device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1744665A (en) * | 2005-09-14 | 2006-03-08 | 上海广电（集团）有限公司中央研究院 | Processing method for point-to-point increasing video image clarity |
| US20080226148A1 (en) * | 2007-03-16 | 2008-09-18 | Sti Medical Systems, Llc | Method of image quality assessment to produce standardized imaging data |
| CN101282481A (en) * | 2008-05-09 | 2008-10-08 | 中国传媒大学 | Method for evaluating video quality based on artificial neural net |
| CN101426150A (en) * | 2008-12-08 | 2009-05-06 | 青岛海信电子产业控股股份有限公司 | Video image quality evaluation method and system |
| CN102033937A (en) * | 2010-12-20 | 2011-04-27 | 百度在线网络技术（北京）有限公司 | Method and system for displaying video search result |
-
2011
- 2011-08-19 CN CN201110240628.5A patent/CN102955947B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1744665A (en) * | 2005-09-14 | 2006-03-08 | 上海广电（集团）有限公司中央研究院 | Processing method for point-to-point increasing video image clarity |
| US20080226148A1 (en) * | 2007-03-16 | 2008-09-18 | Sti Medical Systems, Llc | Method of image quality assessment to produce standardized imaging data |
| CN101282481A (en) * | 2008-05-09 | 2008-10-08 | 中国传媒大学 | Method for evaluating video quality based on artificial neural net |
| CN101426150A (en) * | 2008-12-08 | 2009-05-06 | 青岛海信电子产业控股股份有限公司 | Video image quality evaluation method and system |
| CN102033937A (en) * | 2010-12-20 | 2011-04-27 | 百度在线网络技术（北京）有限公司 | Method and system for displaying video search result |
Non-Patent Citations (1)
| Title |
|---|
| 刘建武: "文档图像的清晰度评价研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106713864A (en) * | 2016-12-24 | 2017-05-24 | 深圳云天励飞技术有限公司 | Video processing method and device |
| CN106713864B (en) * | 2016-12-24 | 2018-09-21 | 深圳云天励飞技术有限公司 | A kind of method for processing video frequency and device |
| WO2019137167A1 (en) * | 2018-01-10 | 2019-07-18 | Oppo广东移动通信有限公司 | Photo album management method and apparatus, storage medium, and electronic device |
| CN109801288A (en) * | 2019-01-25 | 2019-05-24 | 淮阴师范学院 | A kind of image Focus field emission array implementation method based on directional statistics characteristic |
| CN110175980A (en) * | 2019-04-11 | 2019-08-27 | 平安科技（深圳）有限公司 | Image definition recognition methods, image definition identification device and terminal device |
| WO2020206912A1 (en) * | 2019-04-11 | 2020-10-15 | 平安科技（深圳）有限公司 | Image definition recognition method, image definition recognition apparatus, and terminal device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102955947B (en) | 2017-09-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106485183B (en) | A kind of Quick Response Code localization method and system | |
| Li et al. | Multi-scale single image dehazing using laplacian and gaussian pyramids | |
| CN106780485A (en) | SAR image change detection based on super-pixel segmentation and feature learning | |
| CA2914043A1 (en) | Evaluating image sharpness | |
| EP2587826A1 (en) | Extraction and association method and system for objects of interest in video | |
| US20140205206A1 (en) | Systems and methods for resizing an image | |
| WO2020248866A1 (en) | Method and system for image search and cropping | |
| CN102968413A (en) | Method and equipment for providing searching result | |
| CN110678878B (en) | Apparent feature description attribute identification method and device | |
| CN101339661B (en) | Real time human-machine interaction method and system based on moving detection of hand held equipment | |
| CN104281864A (en) | Method and equipment for generating two-dimensional codes | |
| CN107945111A (en) | A kind of image split-joint method based on SURF feature extraction combination CS LBP descriptors | |
| JP4772819B2 (en) | Image search apparatus and image search method | |
| CN108491786A (en) | A kind of method for detecting human face based on hierarchical network and Cluster merging | |
| CN102955947A (en) | Equipment and method for determining image definition | |
| CN103995864B (en) | A kind of image search method and device | |
| CN106815588A (en) | Rubbish image filtering method and device | |
| CN108304839A (en) | A kind of image processing method and device | |
| CN103810488A (en) | Image feature extraction method, image searching method, server, terminal and system | |
| CN110120059B (en) | Icon generation method and device | |
| CN104615639A (en) | Method and equipment for providing showing information of picture | |
| JP6363547B2 (en) | Information processing apparatus and sentence imaging program | |
| JP6838167B2 (en) | Web page main image recognition method and device | |
| CN104133845A (en) | Multimedia information display method and multimedia information display processing equipment | |
| CN105828061B (en) | A kind of virtual view quality evaluating method of view-based access control model masking effect |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20130306 Assignee: Beijing small mutual Entertainment Technology Co., Ltd. Assignor: BEIJING BAIDU NETCOM SCIENCE AND TECHNOLOGY Co.,Ltd. Contract record no.: 2017110000013 Denomination of invention: Equipment and method for determining image definition License type: Exclusive License Record date: 20170705 |
|
| EE01 | Entry into force of recordation of patent licensing contract | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |