CN106095876A

CN106095876A - Image processing method and device

Info

Publication number: CN106095876A
Application number: CN201610395671.1A
Authority: CN
Inventors: 陈志军; 王百超; 杨松
Original assignee: Beijing Xiaomi Mobile Software Co Ltd
Current assignee: Beijing Xiaomi Mobile Software Co Ltd
Priority date: 2016-06-06
Filing date: 2016-06-06
Publication date: 2016-11-09
Anticipated expiration: 2036-06-06
Also published as: CN106095876B

Abstract

The disclosure is directed to a kind of image processing method and device, described method includes: obtain the first image from default destination object photograph album；Image in described destination object photograph album all comprises fisrt feature；In described first image, the characteristic area comprising second feature is selected based on described fisrt feature；Area characteristic information is extracted for described characteristic area；The second image is searched for based in described area characteristic information image beyond described destination object photograph album；Described second image comprises the feature similar to described second feature.This image processing techniques makes user search for the image relevant to destination object photograph album without artificial traversal all images, just can automatically search and belong to the outer image of destination object photograph album of same destination object with the image in destination object photograph album, thus improve accuracy and the convenience that image arranges, bring great convenience to the use of user.

Description

Image processing method and device

Technical field

It relates to technical field of image processing, particularly relate to image processing method and device.

Background technology

Along with the development of ICT technology, Cloud Server and terminal unit is relied on to be analyzed data and manage just becoming more Come the most universal.Such as, shoot photo with mobile terminal, utilize the cloud photograph album on Cloud Server and the phase on terminal unit afterwards Volume carries out taxonomic revision to photo.

Although relevant technology can be by pre-conditioned, such as face, taxonomic revision goes out photograph album belonging to subject, but Be, under many circumstances, subject do not meet the image that pre-conditioned image, such as face are blocked, it is impossible to quilt Automatically identify thus cannot be classified in corresponding photograph album.Now, though the image outside corresponding photograph album and the figure in corresponding photograph album As having the common ground beyond pre-conditioned, also cannot be by automatic clustering to corresponding photograph album.Now, in addition it is also necessary to traveled through entirely by user Portion's image, and therefrom select the image wanted and shifting is stored in corresponding photograph album, this way wastes time and energy, and brings to user Greatly inconvenience.

Summary of the invention

For overcoming problem present in correlation technique, disclosure embodiment provides image processing method and device.

First aspect according to disclosure embodiment, it is provided that a kind of image processing method, described method includes:

The first image is obtained from default destination object photograph album；Image in described destination object photograph album all comprises first Feature；

In described first image, the characteristic area comprising second feature is selected based on described fisrt feature；

Area characteristic information is extracted for described characteristic area；

The second image is searched for based in described area characteristic information image beyond described destination object photograph album；Described Two images comprise the feature similar to described second feature.

Alternatively, described method also includes:

Described second image is moved into described destination object photograph album.

Alternatively, the second figure is searched for based in described area characteristic information image beyond described destination object photograph album Picture, including:

Based on described area characteristic information, the mode of sliding window is used to judge the image beyond described destination object photograph album Whether comprise the feature similar to described second feature；And

When judging that the piece image beyond described destination object photograph album comprises the feature similar to described second feature, Determine that described piece image is described second image.

Candidate region is selected on image beyond described destination object photograph album；

Candidate feature information is extracted for described candidate region；

Judge that whether the described candidate feature information similarity with described area characteristic information is more than threshold value；And

When the similarity of described candidate feature information with described area characteristic information is more than described threshold value, determine that there is institute The image stating candidate region comprises the feature similar to described second feature.

Alternatively, described fisrt feature is the face of destination object.

Alternatively, described method also includes:

When described first image comprises the face of many individuals, for the plurality of people of facial recognition of the plurality of people Age information and/or gender information；And

Age information according to the plurality of people identified and/or gender information determine the face of described destination object Hole.

Alternatively, described second feature is the clothes of destination object.

Alternatively, described method also includes:

According to the attribute of image, all images within described destination object photograph album and in addition is grouped；

Judge to be grouped in each group image of gained the image whether existed within described destination object photograph album；And

When judging one group of image exists the image within described destination object photograph album, from being both present in described one group Image is present in again in the image within described destination object photograph album and obtains the first image.

Alternatively, during described second image is present in described one group of image.

Alternatively, the attribute of described image includes the shooting time of image, picture size, image resolution ratio, video bits Size, image file format at least one.

Alternatively, described characteristic area comprises described fisrt feature and described second feature.

Alternatively, described characteristic area is rectangular area.

Alternatively, extract area characteristic information for described characteristic area, including:

Described characteristic area is normalized to a fixed size, then extracts region for described characteristic area Characteristic information.

Alternatively, described candidate region is rectangular area.

Alternatively, extract candidate feature information for described candidate region, including:

Described candidate region is normalized to a fixed size, then extracts candidate for described candidate region Characteristic information.

Alternatively, described method also includes:

Described second image is moved in other photograph album being stored to beyond described destination object photograph album.

Alternatively, described method also includes:

Described second image is chosen or delete processing.

Second aspect according to disclosure embodiment, it is provided that a kind of image processing apparatus, described device includes:

Acquisition module, is configured to obtain the first image from default destination object photograph album；Described destination object photograph album In image all comprise fisrt feature；

Select module, be configured to select to comprise second feature in described first image based on described fisrt feature Characteristic area；

Extraction module, is configured to for described characteristic area to extract area characteristic information；

Search module, is configured to search based in described area characteristic information image beyond described destination object photograph album Rope the second image；Described second image comprises the feature similar to described second feature.

Alternatively, described device also includes:

Mobile module, is configured to described second image is moved into described destination object photograph album.

Alternatively, described search module includes:

First judges submodule, is configured to, based on described area characteristic information, use the mode of sliding window to judge institute State whether the image beyond destination object photograph album comprises the feature similar to described second feature；And

Second judges submodule, is configured as described first and judges that submodule is judged beyond described destination object photograph album Piece image when comprising the feature similar to described second feature, determine that described piece image is described second image.

Alternatively, described search module includes:

First selects submodule, is configured on the image beyond described destination object photograph album select candidate region；

First extracts submodule, is configured to for described candidate region to extract candidate feature information；

3rd judges submodule, is configured to the similarity judging described candidate feature information with described area characteristic information Whether more than threshold value；And

4th judges submodule, is configured as the described 3rd and judges that submodule judges described candidate feature information and institute When stating the similarity of area characteristic information more than described threshold value, determine that the image with described candidate region comprises and described second The feature of feature similarity.

Alternatively, described fisrt feature is the face of destination object.

Alternatively, described device also includes:

Identification module, when being configured as in described first image the face comprising many individuals, for the plurality of people's The age information of the plurality of people of facial recognition and/or gender information；And

First judge module, be configured to the age information of the plurality of people that identifies according to described identification module and/ Or gender information determines the face of described destination object.

Alternatively, described second feature is the clothes of destination object.

Alternatively, described device also includes:

Grouping module, is configured to the attribute according to image to all images within described destination object photograph album and in addition It is grouped；And

Second judge module, is configured to judge to be grouped in each group image of gained whether there is described destination object photograph album Within image；

Wherein, described acquisition module is configured as described second judge module and judges to there is described mesh in one group of image During image within mark object photograph album, within being not only present in described one group of image but also be present in described destination object photograph album Image obtains the first image.

Alternatively, described characteristic area is rectangular area.

Alternatively, described extraction module is configured to:

Alternatively, described candidate region is rectangular area.

Alternatively, described first extraction submodule is configured to:

Alternatively, described device also includes:

Mobile module, is configured to described second image is moved other photograph album being stored to beyond described destination object photograph album In.

Alternatively, described device also includes:

Processing module, is configured to choose described second image or delete processing.

The third aspect according to disclosure embodiment, it is provided that a kind of image processing apparatus, described device includes:

Processor；

For storing the memorizer of processor executable；

Wherein, described processor is configured to:

Embodiment of the disclosure that the technical scheme of offer can include following beneficial effect:

Disclosure embodiment provides a kind of image processing techniques, when the destination object photograph album of user is polymerized by fisrt feature During the first image, this image processing techniques can select characteristic area comprising second feature based on fisrt feature and carry Take area characteristic information such that it is able to search for beyond photograph album based on this area characteristic information and there is spy similar with second feature The image levied, this makes it possible to the image farthest avoiding omitting destination object.Therefore, this image processing techniques makes to use The image relevant to destination object photograph album is searched for without artificial traversal all images in family, it becomes possible to automatically search and target pair As the image in photograph album belongs to the outer image of destination object photograph album of same destination object, thus improve the accuracy that image arranges And convenience, bring great convenience to the use of user.

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 for explaining the principle of the disclosure together with description.

Fig. 1 is the flow chart according to a kind of image processing method shown in the disclosure one exemplary embodiment；

Fig. 2 is the design sketch for the application scenarios according to the disclosure one specific embodiment is described；

Fig. 3 is according to a kind of implementing procedure figure of step S140 in the Fig. 1 shown in disclosure another exemplary embodiment；

Fig. 4 is according to a kind of implementing procedure figure of step S140 in the Fig. 1 shown in disclosure another exemplary embodiment；

Fig. 5 is to determine target pair in multiple faces in the image processing method according to disclosure another exemplary embodiment One implementing procedure of the face of elephant；

Fig. 6 is the enforcement stream obtaining the first image in the image processing method according to disclosure another exemplary embodiment Journey；

Fig. 7 is another flow chart being embodied as a kind of image processing method exemplified according to the disclosure；

Fig. 8 is the block diagram according to a kind of image processing apparatus shown in the disclosure one exemplary embodiment；

Fig. 9 is the block diagram of the structure according to the search module 840 in the Fig. 8 shown in the disclosure one exemplary embodiment；

Figure 10 is the block diagram of the structure according to the search module 840 in the Fig. 8 shown in disclosure another exemplary embodiment；

Figure 11 be according in a kind of image processing apparatus shown in disclosure another exemplary embodiment in multiple faces Determine the block diagram of the part of the face of destination object；

Figure 12 is acquisition the first image according to a kind of image processing apparatus shown in disclosure another exemplary embodiment The block diagram of part；

Figure 13 is the block diagram according to a kind of image processing apparatus shown in the disclosure one exemplary embodiment；

Figure 14 is the block diagram according to the another kind of image processing apparatus shown in the disclosure one exemplary embodiment.

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.

Fig. 1 is the flow chart according to a kind of image processing method shown in the disclosure one exemplary embodiment, the method bag Include following steps S110-S140:

In step s 110, from default destination object photograph album, the first image is obtained；Image in destination object photograph album All comprise fisrt feature.

In one embodiment, fisrt feature can be the face of destination object.In the case, the destination object preset Photograph album can be to use facial recognition and face clustering method, is placed on by the photo in the photo of user with same person face Same photograph album is formed.

Fig. 5 is to determine target pair in multiple faces in the image processing method according to disclosure another exemplary embodiment One implementing procedure of the face of elephant.

As it is shown in figure 5, in one embodiment, also may be used according to the image processing method of disclosure another exemplary embodiment To include step S510 and S520.In step S510, when the first image comprises the face of many individuals, for the plurality of The age information of the plurality of people of facial recognition of people and/or gender information.In step S520, described many according to identify The age information of individual and/or gender information determine the face of destination object.By the flow process shown in Fig. 5, according to the disclosure The image processing method of another exemplary embodiment can determine mesh rapidly and accurately in the image of the face comprising many individuals The face of mark object, i.e. fisrt feature.

In the step s 120, in the first image, the characteristic area comprising second feature is selected based on fisrt feature.

In one embodiment, second feature is the clothes of destination object.In this case, it is possible to based on fisrt feature, The face of such as destination object, selects the characteristic area of a clothes comprising destination object in the first image.

In one embodiment, characteristic area comprises fisrt feature and second feature.In another embodiment, characteristic area Territory can not comprise fisrt feature, and comprises second feature.Such as, it is the face of destination object and second feature when fisrt feature When being the clothes of destination object, characteristic area comprises the clothes of destination object, and this is owing to destination object photograph album is based on target The face of object and set up, the image outside destination object photograph album may not comprise the face of destination object, and may bag Clothes containing destination object.In other words, characteristic area comprises second feature so that in the method according to disclosure embodiment, Can search for and comprise and the image of second feature similar characteristics.

In step s 130, area characteristic information is extracted for characteristic area.

In one embodiment, characteristic area is rectangular area.In some cases, rectangular area is conducive to carrying out feature The operations such as extraction.In another embodiment, other shapes of region, such as, polygonal beyond circular, oval or rectangle Characteristic area is also feasible.In other words, as long as feature information extraction can be carried out, the characteristic area of any shape is all feasible 's.

In one embodiment, step S130 includes: characteristic area normalizes to fixed size, then a pin Characteristic area is extracted area characteristic information.Normalization makes image can resist the attack of geometric transformation, and can find out Those invariants in image.Therefore, after normalizing to fixed size, extract the characteristic information of characteristic area again, make Such characteristic information search graph picture obtained must be utilized more accurate.

In step S140, search for the second image based in area characteristic information image beyond destination object photograph album； Second image comprises the feature similar to second feature.

Above-mentioned image processing method as is shown in fig. 1 is further illustrated referring to the application scenarios shown in Fig. 2.

Fig. 2 is the design sketch for the application scenarios according to the disclosure one specific embodiment is described.In fig. 2, target pair As photograph album 200 comprises the image 220 and 230 of destination object 210, do not comprise image 240.It addition, the destination object phase shown in Fig. 2 It is only example that volume 200 comprises two width images 220 and 230, and destination object photograph album 200 can comprise piece image, it is also possible to bag Containing three width or more multiple image.

When by shown in Fig. 1 above-mentioned image processing method be applied to shown in Fig. 2 application scenarios time, in step S110 In, from default destination object photograph album 200, obtain the first image 220；Image 220 and 230 in destination object photograph album 200 is equal Comprise fisrt feature 221, the i.e. face of destination object 210.According to can to the explanation of above-mentioned image processing method above by reference to Fig. 1 Know, when utilizing the face cohesive image of destination object 210 in destination object photograph album 200, the face of destination object 210 not by The image 240 photographed is not aggregated in destination object photograph album 200.

In the step s 120, select to comprise second feature 222, i.e. mesh in the first image 220 based on fisrt feature 221 Mark object 210 clothes, characteristic area 225.According to above by reference to the Fig. 1 explanation to above-mentioned image processing method, though So the characteristic area 225 shown in Fig. 2 comprises both fisrt feature 221 and second feature 222, but, characteristic area 225 can wrap Fisrt feature 221 is not comprised containing second feature 222.Although the image 220 that it should be noted that in Fig. 2 is identical with 230, but Be this be only example, image 220 and 230 can have identical fisrt feature 221, without identical.

In step s 130, area characteristic information is extracted for characteristic area 225.In the embodiment shown in Figure 2, special Levying region 225 is rectangular area, it will be appreciated that the disclosure is not limited to this.

In step S140, search for the second figure based in area characteristic information image beyond destination object photograph album 200 As 240；Second image 240 comprises the feature 242 similar to second feature 222.

Knowable to the above-mentioned image processing method as is shown in fig. 1 with reference to the application scenarios explanation shown in Fig. 2, work as user Destination object photograph album 200 when being polymerized the first image by fisrt feature 221, this image processing method can be based on fisrt feature 221 select a characteristic area comprising second feature 222 225 and extract area characteristic information such that it is able to based on this region Characteristic information search beyond destination object photograph album 200 has and second feature 222, this makes it possible to farthest avoid Omit the image of destination object.Therefore, this image processing method makes user search for and mesh without artificial traversal all images The image that mark object photograph album is relevant, it becomes possible to automatically search and belong to same destination object with the image in destination object photograph album The outer image of destination object photograph album, thus improve accuracy and the convenience that image arranges, bring greatly to the use of user Convenience.

Although it should be noted that the image processing method above by reference to Fig. 1 and Fig. 2 description can be by the first image 220 The face of destination object is as fisrt feature 221, and using the clothes of destination object as second feature 222, but the disclosure is not It is limited to this.Such as, fisrt feature can be the clothes 222 of destination object, and second feature is the face 221 of destination object.Again Such as, fisrt feature is the clothes 222 of destination object, and the medicated cap (not shown) that second feature is destination object to be worn.Again Such as, fisrt feature is the face of destination object, and the second feature hand-held bouquet that is destination object or balloon (do not show in figure Go out).In other words, fisrt feature and second feature can be the features that in image, any two differs.

Referring to other implementing procedure of step S140 in Fig. 3 and Fig. 4 explanatory diagram 1.

Fig. 3 is according to a kind of implementing procedure figure of step S140 in the Fig. 1 shown in disclosure another exemplary embodiment.As Shown in Fig. 3, step S140 in Fig. 1 can also include step S1401 and S1402.

In step S1401, based on area characteristic information, the mode of sliding window is used to judge beyond destination object photograph album Image whether comprise the feature similar to second feature.

In step S1402, similar to second feature when judging that the piece image beyond destination object photograph album comprises During feature, determine that described piece image is the second image.

Referring to step S1401 illustrated in the application scenarios explanatory diagram 3 of Fig. 2 and S1402.

In step S1401, based on area characteristic information, use sliding window (such as, the sliding window shown in Fig. 2 245) mode judges whether the image beyond destination object photograph album comprises the feature similar to second feature 222.

In step S1402, similar to second feature 222 when judging that the piece image beyond destination object photograph album comprises Feature 242 time, determine that image 240 is the second image.

Knowable to an implementing procedure of step S140 as shown in Figure 3 with reference to the application scenarios explanation shown in Fig. 2, when Based on area characteristic information when search has with second feature 222 destination object photograph album 200 in addition, can be sliding by using Dynamic window judges whether the feature 242 similar with second feature 222 in the picture, this makes it possible to avoid at a width figure Omit possible similar characteristics 242 in Xiang, and then can farthest avoid omitting the image of destination object.Therefore, this figure As treatment technology makes user search for the image relevant to destination object photograph album without artificial traversal all images, it becomes possible to from Move and search image outside the destination object photograph album belonging to same destination object with the image in destination object photograph album, thus improve The accuracy of image arrangement and convenience, bring great convenience to the use of user.

Fig. 4 is according to a kind of implementing procedure figure of step S140 in the Fig. 1 shown in disclosure another exemplary embodiment.As Shown in Fig. 4, step S140 in Fig. 1 can also include step S1403-S1406.

In step S1403, the image beyond destination object photograph album selects candidate region.

In step S1404, extract candidate feature information for candidate region.

In one embodiment, candidate region is rectangular area.In some cases, rectangular area is conducive to carrying out feature The operations such as extraction.In another embodiment, other shapes of region, such as, polygonal beyond circular, oval or rectangle Candidate region is also feasible.In other words, as long as feature information extraction can be carried out, the candidate region of any shape is all feasible 's.

In one embodiment, step S1404 includes: candidate region normalizes to a fixed size, then Candidate feature information is extracted for candidate region.Normalization makes image can resist the attack of geometric transformation, and can look for Publish picture as in those invariants.Therefore, after normalizing to fixed size, extract the candidate feature of candidate region again Information so that utilize such candidate feature information search image obtained more accurate.

In step S1405, it is judged that whether candidate feature information is more than threshold value with the similarity of area characteristic information.Threshold value Can determine according to great many of experiments and analysis, not do specific restriction at this.

In step S1406, when the similarity of candidate feature information with area characteristic information is more than threshold value, determines and have The image of candidate region comprises the feature similar to second feature.

Referring to step S1403-S1406 illustrated in the application scenarios explanatory diagram 4 of Fig. 2.

In step S1403, the image 240 beyond destination object photograph album 200 selects candidate region 245.

In step S1404, extract candidate feature information for candidate region 245.

In step S1405, it is judged that whether candidate feature information is more than threshold value with the similarity of area characteristic information.

In step S1406, when the similarity of candidate feature information with area characteristic information is more than threshold value, determines and have The image 240 of candidate region comprises the feature 242 similar to second feature 222.

Knowable to an implementing procedure of step S140 as shown in Figure 4 with reference to the application scenarios explanation shown in Fig. 2, when Based on area characteristic information when search has with second feature 222 destination object photograph album 200 in addition, can exist by using Select candidate feature region 245 on image beyond destination object photograph album and extract candidate feature letter for candidate region 245 Breath, this makes it possible to come accurate judging characteristic region 225 by the similarity of comparison domain characteristic information and candidate feature information With the similarity in candidate feature region 245, and then can judge whether have on the image beyond destination object photograph album exactly There is the feature 242 similar with second feature 222.So, it becomes possible to judge that piece image is the second image exactly.Cause This, this image processing techniques makes user search for the image relevant to destination object photograph album without artificial traversal all images, Just can automatically search and belong to the outer image of destination object photograph album of same destination object with the image in destination object photograph album, from And improve accuracy and the convenience that image arranges, bring great convenience to the use of user.

Illustrate to obtain the first image in the image processing method of disclosure another exemplary embodiment referring to Fig. 6 An implementing procedure.

Fig. 6 is the enforcement stream obtaining the first image in the image processing method according to disclosure another exemplary embodiment Journey.

As shown in Figure 6, step S610-is also included according to the image processing method of disclosure another exemplary embodiment S630。

In step S610, all images within destination object photograph album and in addition carried out point according to the attribute of image Group.

In step S620, it is judged that whether each group image of packet gained exists the image within destination object photograph album.

In step S630, when judging one group of image exists the image within destination object photograph album, from both existing It is present in again in described one group of image in the image within destination object photograph album and obtains the first image.

In one embodiment, during the second image is present in described one group of image.In other words, the second image had both been present in institute State in one group of image, be present in again outside described destination object photograph album.

In one embodiment, the attribute of image includes the shooting time of image, picture size, image resolution ratio, image Bit size, image file format at least one.

By all images within destination object photograph album and in addition is grouped by the attribute according to image, this enforcement The image processing method of example can either improve the probability searching the second image, can reduce again the scope of image to be searched, The speed of search the second image can also be improved.

Such as, within target image photograph album, there are 40 width images, beyond target image photograph album, have 60 width images, totally 100 width figure Picture.Utilize according to the image processing method of disclosure another exemplary embodiment, in step S610, during according to the shooting of image Between section this 100 width image is divided into such as 5 time periods, i.e. 5 groups, often group amount of images may equal be likely to.In step In rapid S620, such as, have selected one group of 20 width image, it is judged that within whether this group 20 width image exists destination object photograph album Image.In step S630, when judging this group 20 width image exists 5 width image within destination object photograph album, from this 5 Width image obtains the first image.Owing in the time period, the image of user's shooting is likely to more than one, and at one In time period, the second feature (such as clothes) of user typically will not change.Therefore, when having 5 first in 20 photos of this group During image, there will be likely some in remaining the 15 width image in this group 20 width image are described and photographed the second of destination object The image of feature (such as clothes).Obviously, search in this set remaining 15 width image the probability of the second image than The probability searching the second image in 60 width images beyond whole destination object photograph albums wants height, and the minimizing of volumes of searches is said Bright search speed is faster.

The most such as, if above-mentioned one group of 20 width image is entirely the image in destination object photograph album, then, according to these public affairs The image processing method opening another exemplary embodiment can not select the first image in this 20 width image, this is because this group In the same time period of 20 width images, there is no an image beyond destination object photograph album, therefore complete beyond destination object photograph album The probability that portion 60 width searches the second image is relatively low.Obviously, in this organizes 20 width images, do not obtain the first image, but from both depositing It is in one group of image to be present in again in the image within destination object photograph album and obtains the first image, can either improve and search The probability of two images, can reduce again the scope of image to be searched, additionally it is possible to improve the speed of search the second image.

Image is grouped according to the shooting time of image although foregoing illustrates only, but it is also possible to according to image chi In the file format of very little, image resolution ratio, video bits size, image at least one image is grouped.By according to figure All images within destination object photograph album and in addition is grouped by the above-mentioned attribute of picture, the image processing method of the present embodiment The probability searching the second image can either be improved, the scope of image to be searched can be reduced again, additionally it is possible to improve search the The speed of two images.

In an embodiment of the disclosure, above-described image processing method can also include step: by the second figure As moving into destination object photograph album.The second image searched is moved into destination object photograph album and can improve the accuracy that image arranges And convenience, bring great convenience to the use of user.

In an embodiment of the disclosure, above-described image processing method can also include step: by the second figure As moving in other photograph album being stored to beyond destination object photograph album.The second image searched is moved and is stored to beyond destination object photograph album Other photograph album in can improve image arrange accuracy and convenience, bring great convenience to the use of user.

In an embodiment of the disclosure, above-described image processing method can also include step: to the second figure As carrying out choosing or delete processing.The second image searched is chosen or delete processing can improve the standard that image arranges Exactness and convenience, bring great convenience to the use of user.

Fig. 7 is another flow chart being embodied as a kind of image processing method exemplified according to the disclosure.

In the image processing method of this specific embodiment, for application-specific scene, i.e. arrange baby's photograph album and (that is, protect Deposit the photograph album of the image of a specific baby/child), carry out whole image processing flow S710-S765.

As it is shown in fig. 7, in step S710, generate baby's photograph album by recognition of face and face cluster.That is, first pass through Baby's photograph album will flock together most baby's image.Such as, initially with recognition of face and face cluster method, will The image of the face comprising same baby in the image of user is placed in same photograph album.Comprise the image of the face of different people Can be placed in different photograph albums.The concrete mode creating baby's photograph album, has and automatically identifies that baby is to create baby's photograph album and use Family manual creation baby's photograph album two kinds.

In step S715, identified by the age and/or baby that sex identification determines in image.After having clustered, Sex and/or age recognition methods, the recognition of face to each photograph album can be used, if the age is less than certain threshold value, then recognize Surely it is baby.

In step S720, by intervals, image is grouped and the image of each packet is analyzed. Owing to baby is taking pictures when, generally not it is in the situation of cooperation, as couched, the figure viewed from behind, tosses about.This examines to face Survey adds difficulty, and the most generally, the photo recall rate of baby is lower than general adult's.But if baby father and mother Words, the figure viewed from behind picture of baby, it is the most also liked.Therefore can use method based on picture search, increase recalling of baby Rate.This is because what the baby's clothes etc. within the same time usually will not become.

In step S725, it is judged that whether packet exists baby's image.

In step S730, when judging packet exists baby's image, the human face region of baby is enlarged, really A fixed rectangular area including baby's face, clothes.When judging packet does not exist baby's image, terminate this image place Reason method.

In step S735, rectangular area is normalized to a fixed size, then to its extracted region Area characteristic information.Such as, area characteristic information can be convolution feature, the face trained by convolutional neural networks (CNN) The characteristic informations such as color textural characteristics.

In step S740, using sliding window, other image beyond baby's photograph album in being grouped this chooses candidate Frame travels through.Such as, during search determines this packet, other photos do not comprise the photo of this baby (in baby's photograph album Have determined that), then use the form of sliding window to choose from top to bottom to every pictures, from left to right, choose different size of Candidate frame travels through.

In step S745, often choose a candidate frame, candidate frame is normalized to a fixed size, so Afterwards its region is carried candidate and take candidate feature information.

In step S750, candidate feature information and area characteristic information are compared, determine candidate feature information with The similarity of area characteristic information.That is, the feature during contrast includes the rectangular area of baby's face, clothes and the spy in candidate frame The similarity levied.

In step S755, it is judged that whether candidate feature information is more than threshold value with the similarity of area characteristic information.

In step S760, when determining the candidate feature information similarity with area characteristic information more than threshold value, explanation Including baby's face, clothes rectangular area in feature and feature similarity in candidate frame, then this candidate frame is sought exactly The baby region looked for, now moves into baby's photograph album by this image.One pictures only one of which target baby in theory, after finding, Then the search of diagram picture terminates.When determining that the candidate feature information similarity with area characteristic information is not more than threshold value, this width The search of image also terminates.

In step S765, it may be judged whether have stepped through whole images.When determine have stepped through whole images time, then A kind of image processing method according to another specific embodiment of the disclosure shown in Fig. 7 terminates.When determine also do not travel through whole During image, return in step S740.Now, this in each packet is searched for according to the method for above-mentioned steps S740-S765 Baby.Then the image searched is joined in baby's photograph album, which enhance the recall rate of baby's photograph album.

According in the another kind of image processing method of another specific embodiment of the disclosure, first passing through baby's photograph album will be Most baby flocks together, and then obtains one rectangular area of frame of each baby in baby's photograph album, then extracts spy Levy, other the photo in search similar time section, such as the figure viewed from behind etc., increase the recall rate of baby's photograph album.

When creating baby's photograph album, can first detect face, then use disaggregated model to judge this face whether Baby's, if being added in baby's photograph album.Owing to tampering when baby is frequent or mismatching, or the face quilt of baby The reason such as blocking, the face of this baby substantially can not be detected or can miss.Such as, 100 faces, Bao Baozheng altogether 60 of face, other block face or the same baby 40 without face, normal recall rate 60%.User is typically taking pictures Time, typically incessantly clap one in the time period, by, after the face that normally detects baby, then being cut out its Garment region Cut.Based on its Garment region, the picture of 40 undetected babies of search.So baby's photograph album recall rate just improves ?.

In one embodiment, if it is known that baby currently what to wear, the figure of baby can not detected from other Baby's face that may be present that in Xiang, removal search is similar.The mode of similar picture search, on concrete methods of realizing, is first face Photograph album is polymerized, and then identifies baby by the age, then intercepts the information of baby Garment region, then adopts from other pictures Go to identify whether to there may be this by the mode of sliding window+picture search (can be CNN or support vector machine (SVM) etc.) Baby.

In one embodiment, identify baby by the age, then intercept the information of baby Garment region, then from it His picture use the mode of sliding window+picture search (can be CNN or SVM etc.) go to identify whether to there may be this treasured Precious.Being equivalent to clothes information is condition, the method then introducing target detection, and the target detection of baby is as baby's target Dynamically change.

Fig. 8 is the block diagram according to a kind of image processing apparatus shown in the disclosure one exemplary embodiment, this image procossing Device includes following acquisition module 810, selects module 820, extraction module 830 and search module 840.

Acquisition module 810 is configured to obtain the first image from default destination object photograph album；In destination object photograph album Image all comprise fisrt feature.

Figure 11 be according in a kind of image processing apparatus shown in disclosure another exemplary embodiment in multiple faces Determine the block diagram of the part of the face of destination object.

As shown in figure 11, in one embodiment, according to the image processing apparatus of disclosure another exemplary embodiment also Identification module 1110 and the first judge module 1120 can be included.Identification module 1110 is configured as comprising in the first image many During the face of individual, for the age information of the plurality of people of facial recognition and/or the gender information of the plurality of people.First sentences Disconnected module 1120 is configured to the age information of the plurality of people that identifies according to identification module 1110 and/or gender information comes Determine the face of destination object.By the block diagram shown in Figure 11, fill according to the image procossing of disclosure another exemplary embodiment Put the face that can determine destination object rapidly and accurately in the image of the face comprising many individuals, i.e. fisrt feature.

Select module 820 to be configured to based on fisrt feature in the first image, select the characteristic area that comprises second feature Territory.

Extraction module 830 is configured to for characteristic area to extract area characteristic information.

In one embodiment, extraction module 830 is configured to characteristic area normalizes to a fixed size, Then area characteristic information is extracted for characteristic area.Normalization makes image can resist the attack of geometric transformation, and energy Enough find out those invariants in image.Therefore, after normalizing to fixed size, extract the feature of characteristic area again Information so that utilize such characteristic information search graph picture obtained more accurate.

Search module 840 is configured to based on searching for second in area characteristic information image beyond destination object photograph album Image；Second image comprises the feature similar to second feature.

The above-mentioned image processing apparatus as shown in Fig. 8 is further illustrated referring to the application scenarios shown in Fig. 2.

When by shown in Fig. 8 above-mentioned image processing apparatus be applied to shown in Fig. 2 application scenarios time, acquisition module 810 It is configured to obtain the first image 220 from default destination object photograph album 200；Image 220 He in destination object photograph album 200 230 all comprise fisrt feature 221, the i.e. face of destination object 210.According to above by reference to Fig. 8 to above-mentioned image processing apparatus Illustrate to understand, when utilizing the face cohesive image of destination object 210 in destination object photograph album 200, the face of destination object 210 The image 240 that Kong Wei is photographed is not aggregated in destination object photograph album 200.

Module 820 is selected to be configured to select to comprise second feature in the first image 220 based on fisrt feature 221 222, i.e. the clothes of destination object 210, characteristic area 225.According to above by reference to the Fig. 8 explanation to above-mentioned image processing apparatus It is recognized that while the characteristic area 225 shown in Fig. 2 comprises both fisrt feature 221 and second feature 222, but, characteristic area 225 can comprise second feature 222 and not comprise fisrt feature 221.Although image 220 He that it should be noted that in Fig. 2 230 is identical, but this is only example, and image 220 and 230 can have identical fisrt feature 221, without complete phase With.

Extraction module 830 is configured to for characteristic area 225 to extract area characteristic information.In the enforcement shown in Fig. 2 In example, characteristic area 225 is rectangular area, it will be appreciated that the disclosure is not limited to this.

Search module 840 is configured to search for based in area characteristic information image beyond destination object photograph album 200 Second image 240；Second image 240 comprises the feature 242 similar to second feature 222.

Knowable to the above-mentioned image processing apparatus as shown in Fig. 8 with reference to the application scenarios explanation shown in Fig. 2, work as user Destination object photograph album 200 when being polymerized the first image by fisrt feature 221, this image processing apparatus can be based on fisrt feature 221 select a characteristic area comprising second feature 222 225 and extract area characteristic information such that it is able to based on this region Characteristic information search beyond destination object photograph album 200 has the image of the feature 242 similar with second feature 222, so Just can farthest avoid omitting the image of destination object.Therefore, this image processing apparatus makes user without artificial time Go through all images and search for the image relevant to destination object photograph album, it becomes possible to automatically search and the figure in destination object photograph album As belonging to the outer image of the destination object photograph album of same destination object, thus improve accuracy and the convenience that image arranges, give The use of user brings great convenience.

Although it should be noted that the image processing apparatus above by reference to Fig. 8 and Fig. 2 description can be by the first image 220 The face of destination object is as fisrt feature 221, and using the clothes of destination object as second feature 222, but the disclosure is not It is limited to this.Such as, fisrt feature can be the clothes 222 of destination object, and second feature is the face 221 of destination object.Again Such as, fisrt feature is the clothes 222 of destination object, and the medicated cap (not shown) that second feature is destination object to be worn.Again Such as, fisrt feature is the face of destination object, and the second feature hand-held bouquet that is destination object or balloon (do not show in figure Go out).In other words, fisrt feature and second feature can be the features that in image, any two differs.

Other structure referring to the search module 840 in Fig. 9 and Figure 10 explanatory diagram 8.

Fig. 9 is the block diagram of the structure according to the search module 840 in the Fig. 8 shown in the disclosure one exemplary embodiment.As Shown in Fig. 9, the search module 840 in Fig. 8 can also include that the first judgement submodule 8401 and second judges submodule 8402.

First judges that submodule 8401 is configured to, based on area characteristic information, use the mode of sliding window to judge target Whether the image beyond object photograph album comprises the feature similar to second feature.

Second judges that submodule 8402 is configured as the first judgement submodule 8401 and judges beyond destination object photograph album Piece image when comprising the feature similar to second feature, determine that described piece image is the second image.

Judge that submodule 8401 and second judges son referring to illustrate first in the application scenarios explanatory diagram 9 of Fig. 2 Module 8402.

First judges that submodule 8401 is configured to, based on area characteristic information, use sliding window (such as, institute in Fig. 2 The sliding window 245 shown) mode judge whether the image beyond destination object photograph album comprises the spy similar to second feature 222 Levy.

Second judges that submodule 8402 is configured as the first judgement submodule 8401 and judges beyond destination object photograph album Piece image when comprising the feature 242 similar to second feature 222, determine that image 240 is the second image.

Knowable to the structure of the search module 840 as shown in Figure 9 with reference to the application scenarios explanation shown in Fig. 2, when based on Area characteristic information search beyond destination object photograph album 200 has the image of the feature 242 similar with second feature 222 Time, the feature 242 similar with second feature 222 can be judged whether in the picture, so by using sliding window Just it can be avoided that omit possible similar characteristics 242 in piece image, and then can farthest avoid omitting target pair The image of elephant.Therefore, this image processing techniques makes user search for and destination object photograph album without artificial traversal all images Relevant image, it becomes possible to automatically search and belong to the destination object phase of same destination object with the image in destination object photograph album The outer image of volume, thus improve accuracy and the convenience that image arranges, bring great convenience to the use of user.

Figure 10 is the block diagram of the structure according to the search module 840 in the Fig. 8 shown in disclosure another exemplary embodiment. As shown in Figure 10, the search module 840 in Fig. 8 can also include the first selection submodule 8403, first extract submodule 8404, 3rd judges that submodule 8405 and the 4th judges submodule 8406.

First selects submodule 8403 to be configured on the image beyond destination object photograph album select candidate region.

First extracts submodule 8404 is configured to for candidate region to extract candidate feature information.

In one embodiment, the first extraction submodule module 8404 is configured to candidate region normalize to one admittedly Sizing size, then extracts candidate feature information for candidate region.Normalization makes image can resist geometric transformation Attack, and those invariants in image can be found out.Therefore, after normalizing to fixed size, candidate is extracted again The candidate feature information in region so that utilize such candidate feature information search image obtained more accurate.

3rd judges whether submodule 8405 is configured to judge the similarity of candidate feature information and area characteristic information More than threshold value.Threshold value can determine according to great many of experiments and analysis, does not do specific restriction at this.

4th judges that submodule 8406 is configured as the 3rd and judges that submodule 8405 judges candidate feature information and district When the similarity of characteristic of field information is more than threshold value, determine that the image with candidate region comprises the feature similar to second feature.

Application scenarios referring to Fig. 2 illustrates that the illustrate first selection submodule 8403, first in Figure 10 extracts son Module the 8404, the 3rd judges that submodule 8405 and the 4th judges submodule 8406.

First selects submodule 8403 to be configured on the image 240 beyond destination object photograph album 200 select candidate regions Territory 245.

First extracts submodule 8404 is configured to for candidate region 245 to extract candidate feature information.

3rd judges whether submodule 8405 is configured to judge the similarity of candidate feature information and area characteristic information More than threshold value.

4th judges that submodule 8406 is configured as the 3rd and judges that submodule 8405 judges candidate feature information and district When the similarity of characteristic of field information is more than threshold value, determine that the image 240 with candidate region comprises similar to second feature 222 Feature 242.

Knowable to the structure of the search module 840 as shown in Figure 10 with reference to the application scenarios explanation shown in Fig. 2, when based on Area characteristic information search beyond destination object photograph album 200 has the image of the feature 242 similar with second feature 222 Time, candidate feature region 245 can be selected on the image beyond destination object photograph album and for candidate region 245 by using Extract candidate feature information, this makes it possible to be come accurately by the similarity of comparison domain characteristic information and candidate feature information Judging characteristic region 225 and the similarity in candidate feature region 245, and then can judge exactly beyond destination object photograph album Image on whether there is the feature 242 similar with second feature 222.So, it becomes possible to judge that piece image is not exactly It it is the second image.Therefore, this image processing techniques makes user search for and destination object phase without artificial traversal all images The image that volume is relevant, it becomes possible to automatically search and belong to the destination object of same destination object with the image in destination object photograph album The outer image of photograph album, thus improve accuracy and the convenience that image arranges, bring great convenience to the use of user.

Illustrate to obtain the first figure in the image processing apparatus of disclosure another exemplary embodiment referring to Figure 12 The part of picture.

Figure 12 is acquisition the first image according to a kind of image processing apparatus shown in disclosure another exemplary embodiment The block diagram of part.

As shown in figure 12, grouping module is also included according to the image processing apparatus of disclosure another exemplary embodiment 1210, the second judge module 1220 and acquisition module 810.

Grouping module 1210 is configured to the attribute according to image to all images within destination object photograph album and in addition It is grouped.

Second judge module 1220 is configured to judge to be grouped in each group image of gained whether there is destination object photograph album Within image.

Acquisition module 810 is configured as the second judge module 1220 and judges to there is destination object photograph album in one group of image Within image time, the image within being not only present in described one group of image but also be present in destination object photograph album obtains first Image.

By all images within destination object photograph album and in addition is grouped by the attribute according to image, this enforcement The image processing apparatus of example can either improve the probability searching the second image, can reduce again the scope of image to be searched, The speed of search the second image can also be improved.

Such as, within target image photograph album, there are 40 width images, beyond target image photograph album, have 60 width images, totally 100 width figure Picture.Utilizing the image processing apparatus according to disclosure another exemplary embodiment, grouping module 1210 is configured to according to image Shooting time section this 100 width image is divided into such as 5 time periods, i.e. 5 groups, often group amount of images may equal be likely to not Deng.Whether the second judge module 1220 is configured to such as have selected one group of 20 width image, it is judged that exist in this group 20 width image Image within destination object photograph album.Acquisition module 810 is configured as the second judge module 1220 and judges this group 20 width figure When there is 5 width image within destination object photograph album in Xiang, from this 5 width image, obtain the first image.Due to a time period The image of interior user shooting is likely to more than one, and the second feature (such as clothes) of user is general in a period of time Will not change.Therefore, when in 20 photos of this group exist 5 the first images time, illustrate in this group 20 width image remaining 15 The image of the second feature (such as clothes) having photographed destination object is there will be likely some in width image.Obviously, in this set Remaining 15 width image in search the probability ratio of the second image in 60 width images beyond whole destination object photograph albums The probability searching the second image wants height, and the minimizing of volumes of searches explanation search speed is faster.

The most such as, if above-mentioned one group of 20 width image is entirely the image in destination object photograph album, then, according to these public affairs The image processing apparatus opening another exemplary embodiment can not select the first image in this 20 width image, this is because this group In the same time period of 20 width images, there is no an image beyond destination object photograph album, therefore complete beyond destination object photograph album The probability that portion 60 width searches the second image is relatively low.Obviously, in this organizes 20 width images, do not obtain the first image, but from both depositing It is in one group of image to be present in again in the image within destination object photograph album and obtains the first image, can either improve and search The probability of two images, can reduce again the scope of image to be searched, additionally it is possible to improve the speed of search the second image.

Although be explained above according to the shooting time of image come to image be grouped, but it is also possible to according to picture size, Image resolution ratio, video bits size, image file format at least one come to image be grouped.By according to image All images within destination object photograph album and in addition is grouped by above-mentioned attribute, and the image processing apparatus of the present embodiment can Enough raisings search the probability of the second image, can reduce again the scope of image to be searched, additionally it is possible to improve search the second figure The speed of picture.

In an embodiment of the disclosure, above-described image processing apparatus can also include that mobile module is (in figure Not shown), it is configured to the second image is moved into destination object photograph album.The second image searched is moved into destination object photograph album Accuracy and convenience that image arranges can be improved, bring great convenience to the use of user.

In an embodiment of the disclosure, above-described image processing apparatus can also include that mobile module is (in figure Not shown), it is configured to move in other photograph album being stored to beyond destination object photograph album the second image.The second figure that will search Accuracy and the convenience that image arranges can be improved, to user's in moving other photograph album being stored to beyond destination object photograph album Use brings great convenience.

In an embodiment of the disclosure, above-described image processing apparatus can also include that processing module (is not shown Go out), it is configured to the second image is chosen or delete processing.The second image searched is chosen or delete processing Accuracy and convenience that image arranges can be improved, bring great convenience to the use of user.

Disclosure embodiment provides a kind of image processing apparatus, including:

Processor；

For storing the memorizer of processor executable；

Wherein, processor is configured to:

The first image is obtained from default destination object photograph album；It is special that image in destination object photograph album all comprises first Levy；

In the first image, the characteristic area comprising second feature is selected based on fisrt feature；

Area characteristic information is extracted for characteristic area；

The second image is searched for based in area characteristic information image beyond destination object photograph album；Second image comprise with The feature that second feature is similar.

Figure 13 is the block diagram according to a kind of image processing apparatus 1300 shown in the disclosure one exemplary embodiment.Such as, Device 1300 can be a client, and this client can be application program, it is also possible to be mobile device, such as mobile phone, meter Calculation machine, digital broadcast terminal, messaging devices, game console, tablet device, armarium, body-building equipment, individual digital Assistant etc..

With reference to Figure 13, device 1300 can include following one or more assembly: processes assembly 1302, memorizer 1304, Power supply module 1306, multimedia groupware 1308, audio-frequency assembly 1310, the interface 1312 of input/output (I/O), sensor cluster 1314, and communications component 1316.

Process assembly 1302 and generally control the integrated operation of device 1300, such as with display, call, data communication, The operation that camera operation and record operation are associated.Process assembly 1302 and can include that one or more processor 1320 performs Instruction, to complete all or part of step of above-mentioned method.Additionally, process assembly 1302 can include one or more mould Block, it is simple to process between assembly 1302 and other assemblies is mutual.Such as, process assembly 1302 and can include multi-media module, With facilitate multimedia groupware 1308 and process between assembly 1302 mutual.

Memorizer 1304 is configured to store various types of data to support the operation at device 1300.These data Example include on device 1300 operation any application program or the instruction of method, contact data, telephone book data, Message, picture, video etc..Memorizer 1304 can by any kind of volatibility or non-volatile memory device or they Combination realizes, such as static RAM (SRAM), Electrically Erasable Read Only Memory (EEPROM), erasable can Program read-only memory (EPROM), programmable read only memory (PROM), read only memory (ROM), magnetic memory, flash memory Reservoir, disk or CD.

The various assemblies that power supply module 1306 is device 1300 provide electric power.Power supply module 1306 can include power management System, one or more power supplys, and other generate, manage and distribute, with for device 1300, the assembly that electric power is associated.

The screen of one output interface of offer that multimedia groupware 1308 is included between described device 1300 and user.? In some embodiments, screen can include liquid crystal display (LCD) and touch panel (TP).If screen includes touch panel, Screen may be implemented as touch screen, to receive the input signal from user.Touch panel includes that one or more touch passes Sensor is with the gesture on sensing touch, slip and touch panel.Described touch sensor can not only sense touch or slide dynamic The border made, but also detect the persistent period relevant to described touch or slide and pressure.In certain embodiments, many Media component 1308 includes a front-facing camera and/or post-positioned pick-up head.When device 1300 is in operator scheme, such as shooting mould When formula or video mode, front-facing camera and/or post-positioned pick-up head can receive the multi-medium data of outside.Each preposition shooting Head and post-positioned pick-up head can be a fixing optical lens system or have focal length and optical zoom ability.

Audio-frequency assembly 1310 is configured to output and/or input audio signal.Such as, audio-frequency assembly 1310 includes a wheat Gram wind (MIC), when device 1300 is in operator scheme, during such as call model, logging mode and speech recognition mode, mike quilt It is configured to receive external audio signal.The audio signal received can be further stored at memorizer 1304 or via communication Assembly 1316 sends.In certain embodiments, audio-frequency assembly 1310 also includes a speaker, is used for exporting audio signal.

I/O interface 1312 provides interface, above-mentioned peripheral interface module for processing between assembly 1302 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 1314 includes one or more sensor, for providing the state of various aspects to comment for device 1300 Estimate.Such as, what sensor cluster 1314 can detect device 1300 opens/closed mode, the relative localization of assembly, such as institute Stating display and keypad that assembly is device 1300, sensor cluster 1314 can also detect device 1300 or device 1,300 1 The position change of individual assembly, the presence or absence that user contacts with device 1300, device 1300 orientation or acceleration/deceleration and dress Put the variations in temperature of 1300.Sensor cluster 1314 can include proximity transducer, is configured to do not having any physics The existence of object near detection during contact.Sensor cluster 1314 can also include optical sensor, as CMOS or ccd image sense Device, for using in imaging applications.In certain embodiments, this sensor cluster 1314 can also include acceleration sensing Device, gyro sensor, Magnetic Sensor, pressure transducer or temperature sensor.

Communications component 1316 is configured to facilitate the communication of wired or wireless mode between device 1300 and other equipment.Dress Put 1300 and can access wireless network based on communication standard, such as WiFi, 2G or 3G, or combinations thereof.Exemplary at one In embodiment, broadcast singal or broadcast that communications component 1316 receives from external broadcasting management system via broadcast channel are relevant Information.In one exemplary embodiment, described communications component 1316 also includes near-field communication (NFC) module, to promote short distance Communication.Such as, can be based on RF identification (RFID) technology in NFC module, Infrared Data Association (IrDA) technology, ultra broadband (UWB) technology, bluetooth (BT) technology and other technologies realize.

In the exemplary embodiment, device 1300 can be by one or more application specific integrated circuits (ASIC), numeral Signal processor (DSP), digital signal processing appts (DSPD), PLD (PLD), field programmable gate array (FPGA), controller, microcontroller, microprocessor or other electronic components realize, be used for performing said method.

In the exemplary embodiment, a kind of non-transitory computer-readable recording medium including instruction, example are additionally provided As included the memorizer 1304 of instruction, above-mentioned instruction can have been performed said method by the processor 1320 of device 1300.Example If, described non-transitory computer-readable recording medium can be ROM, random access memory (RAM), CD-ROM, tape, soft Dish and optical data storage devices etc..

A kind of non-transitory computer-readable recording medium, when the instruction in described storage medium is by the process of mobile terminal When device performs so that mobile terminal is able to carry out a kind of image processing method, and the method includes:

Figure 14 is the block diagram according to the another kind of image processing apparatus shown in the disclosure one exemplary embodiment.Such as, dress Put 1400 and may be provided in a server.With reference to Figure 14, device 1400 includes processing assembly 1422, and it farther includes one Or multiple processor, and by the memory resource representated by memorizer 1432, can be performed by processing assembly 1422 for storage Instruction, such as application program.In memorizer 1432 storage application program can include one or more each Module corresponding to one group of instruction.It is configured to perform instruction, to perform said method additionally, process assembly 1422.

Device 1400 can also include a power supply module 1426 be configured to perform device 1400 power management, one Wired or wireless network interface 1450 is configured to be connected to device 1400 network, and input and output (I/O) interface 1458.Device 1400 can operate based on the operating system being stored in memorizer 1432, such as Windows ServerTM, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM or similar.

A kind of non-transitory computer-readable recording medium, when the instruction in described storage medium is by the processor of server During execution so that server is able to carry out a kind of image processing method, and the method includes:

Those skilled in the art, after considering description and putting into practice disclosure disclosed herein, will readily occur to its of the disclosure Its embodiment.The application 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

1. an image processing method, it is characterised in that including:

The first image is obtained from default destination object photograph album；It is special that image in described destination object photograph album all comprises first Levy；

The second image is searched for based in described area characteristic information image beyond described destination object photograph album；Described second figure As comprising the feature similar to described second feature.

Method the most according to claim 1, it is characterised in that also include:

Method the most according to claim 1, it is characterised in that based on described area characteristic information in described destination object phase Image beyond Ce is searched for the second image, including:

Based on described area characteristic information, use whether the mode of sliding window judges the image beyond described destination object photograph album Comprise the feature similar to described second feature；And

When judging that the piece image beyond described destination object photograph album comprises the feature similar to described second feature, determine Described piece image is described second image.

Candidate feature information is extracted for described candidate region；

When the similarity of described candidate feature information with described area characteristic information is more than described threshold value, determine that there is described time The image of favored area comprises the feature similar to described second feature.

Method the most according to claim 1, it is characterised in that described fisrt feature is the face of destination object.

Method the most according to claim 5, it is characterised in that also include:

When described first image comprises the face of many individuals, for the year of the plurality of people of facial recognition of the plurality of people Age information and/or gender information；And

Age information according to the plurality of people identified and/or gender information determine the face of described destination object.

Method the most according to claim 1, it is characterised in that described second feature is the clothes of destination object.

Method the most according to claim 1, it is characterised in that also include:

When judging one group of image exists the image within described destination object photograph album, from being both present in described one group of image In be present in again in the image within described destination object photograph album and obtain the first image.

Method the most according to claim 8, it is characterised in that described second image is present in described one group of image.

Method the most according to claim 8, it is characterised in that the attribute of described image includes the shooting time of image, figure As size, image resolution ratio, video bits size, image file format at least one.

11. methods according to claim 1, it is characterised in that described characteristic area comprises described fisrt feature and described Second feature.

12. methods according to claim 1, it is characterised in that described characteristic area is rectangular area.

13. methods according to claim 1, it is characterised in that extract area characteristic information for described characteristic area, Including:

Described characteristic area is normalized to a fixed size, then extracts provincial characteristics for described characteristic area Information.

14. methods according to claim 4, it is characterised in that described candidate region is rectangular area.

15. methods according to claim 4, it is characterised in that extract candidate feature information for described candidate region, Including:

Described candidate region is normalized to a fixed size, then extracts candidate feature for described candidate region Information.

16. methods according to claim 1, it is characterised in that also include:

17. methods according to claim 1, it is characterised in that also include:

Described second image is chosen or delete processing.

18. 1 kinds of image processing apparatus, it is characterised in that including:

Acquisition module, is configured to obtain the first image from default destination object photograph album；In described destination object photograph album Image all comprises fisrt feature；

Select module, be configured in described first image, select, based on described fisrt feature, comprise second feature feature Region；

Search module, is configured to based on search in described area characteristic information image beyond described destination object photograph album the Two images；Described second image comprises the feature similar to described second feature.

19. devices according to claim 18, it is characterised in that also include:

20. devices according to claim 18, it is characterised in that described search module includes:

First judges submodule, is configured to, based on described area characteristic information, use the mode of sliding window to judge described mesh Whether the image beyond mark object photograph album comprises the feature similar to described second feature；And

Second judges submodule, is configured as described first and judges that submodule judges beyond described destination object photograph album When width image comprises the feature similar to described second feature, determine that described piece image is described second image.

21. devices according to claim 18, it is characterised in that described search module includes:

3rd judges submodule, whether is configured to judge the similarity of described candidate feature information and described area characteristic information More than threshold value；And

4th judges submodule, is configured as the described 3rd and judges that submodule judges described candidate feature information and described district When the similarity of characteristic of field information is more than described threshold value, determine that the image with described candidate region comprises and described second feature Similar feature.

22. devices according to claim 18, it is characterised in that described fisrt feature is the face of destination object.

23. devices according to claim 22, it is characterised in that also include:

Identification module, when being configured as in described first image the face comprising many individuals, for the face of the plurality of people Identify age information and/or the gender information of the plurality of people；And

First judge module, is configured to the age information of the plurality of people and/or the property identified according to described identification module Other information determines the face of described destination object.

24. devices according to claim 18, it is characterised in that described second feature is the clothes of destination object.

25. devices according to claim 18, it is characterised in that also include:

Grouping module, is configured to the attribute according to image and carries out all images within described destination object photograph album and in addition Packet；And

Second judge module, within being configured to judge be grouped and whether there is described destination object photograph album in each group image of gained Image；

Wherein, described acquisition module is configured as described second judge module and judges to there is described target pair in one group of image Image during as image within photograph album, within being not only present in described one group of image but also be present in described destination object photograph album Middle acquisition the first image.

26. devices according to claim 25, it is characterised in that described second image is present in described one group of image.

27. devices according to claim 25, it is characterised in that the attribute of described image include the shooting time of image, Picture size, image resolution ratio, video bits size, image file format at least one.

28. devices according to claim 18, it is characterised in that described characteristic area comprises described fisrt feature and described Second feature.

29. devices according to claim 18, it is characterised in that described characteristic area is rectangular area.

30. devices according to claim 18, it is characterised in that described extraction module is configured to:

31. devices according to claim 21, it is characterised in that described candidate region is rectangular area.

32. devices according to claim 21, it is characterised in that described first extracts submodule is configured to:

33. devices according to claim 18, it is characterised in that also include:

Mobile module, is configured to move in other photograph album being stored to beyond described destination object photograph album described second image.

34. devices according to claim 18, it is characterised in that also include:

35. 1 kinds of image processing apparatus, it is characterised in that including:

Processor；

For storing the memorizer of processor executable；

Wherein, described processor is configured to:

The second image is searched for based in described area characteristic information image beyond described destination object photograph album；Described second Image comprises the feature similar to described second feature.