CN106095876B

CN106095876B - Image processing method and device

Info

Publication number: CN106095876B
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: 2021-11-09
Anticipated expiration: 2036-06-06
Also published as: CN106095876A

Abstract

The present disclosure relates to an image processing method and apparatus, the method comprising: acquiring a first image from a preset target object photo album; images in the target object photo album all comprise first characteristics; selecting a feature region containing a second feature in the first image based on the first feature; extracting region feature information for the feature region; searching for a second image in images other than the target object album based on the region feature information; the second image includes features similar to the second features. The image processing technology enables a user to automatically search the images outside the target object photo album, which belong to the same target object with the images in the target object photo album, without manually traversing all the images to search the images related to the target object photo album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

Description

Image processing method and device

Technical Field

The present disclosure relates to the field of image processing technologies, and in particular, to an image processing method and apparatus.

Background

With the development of ICT technology, data analysis and management by means of cloud servers and terminal devices are becoming more and more popular. For example, a photo is taken by a mobile terminal, and then the photo is sorted by using a cloud photo album on a cloud server and a photo album on a terminal device.

Although the related art can sort out the albums to which the photographed objects belong according to preset conditions, such as faces, in many cases, images of the photographed objects which do not meet the preset conditions, such as images with blocked faces, cannot be automatically identified and cannot be classified into the corresponding albums. At this time, even if the images outside the corresponding album and the images inside the corresponding album have common points other than the preset conditions, the images cannot be automatically classified into the corresponding album. At this time, the user needs to traverse all the images, select a desired image from the images and transfer the selected image to a corresponding album, which is time-consuming and labor-consuming and brings great inconvenience to the user.

Disclosure of Invention

To overcome the problems in the related art, embodiments of the present disclosure provide an image processing method and apparatus.

According to a first aspect of embodiments of the present disclosure, there is provided an image processing method, the method including:

acquiring a first image from a preset target object photo album; images in the target object photo album all comprise first characteristics;

selecting a feature region containing a second feature in the first image based on the first feature;

extracting region feature information for the feature region;

searching for a second image in images other than the target object album based on the region feature information; the second image includes features similar to the second features.

Optionally, the method further comprises:

and moving the second image into the target object album.

Optionally, searching for a second image in an image other than the target object album based on the region feature information includes:

judging whether the images except the target object album contain the characteristics similar to the second characteristics or not in a sliding window mode based on the regional characteristic information; and

and when judging that an image except the target object album contains the characteristic similar to the second characteristic, determining that the image is the second image.

selecting a candidate area on an image other than the target object album;

extracting candidate feature information for the candidate region;

judging whether the similarity between the candidate characteristic information and the region characteristic information is greater than a threshold value; and

when the similarity between the candidate feature information and the region feature information is greater than the threshold, determining that the image with the candidate region contains a feature similar to the second feature.

Optionally, the first feature is a face of the target object.

Optionally, the method further comprises:

identifying age information and/or gender information of a plurality of persons with respect to faces of the plurality of persons when the faces of the plurality of persons are included in the first image; and

determining a face of the target object according to the identified age information and/or gender information of the plurality of persons.

Optionally, the second characteristic is clothing of the target object.

Optionally, the method further comprises:

grouping all images inside and outside the target object album according to the attributes of the images;

judging whether images within the target object album exist in each group of images obtained by grouping; and

when it is determined that an image within the target object album exists in a group of images, a first image is acquired from images existing both in the group of images and within the target object album.

Optionally, the second image is present in the set of images.

Optionally, the attribute of the image includes at least one of a photographing time of the image, an image size, an image resolution, an image bit size, and a file format of the image.

Optionally, the feature region comprises the first feature and the second feature.

Optionally, the feature region is a rectangular region.

Optionally, extracting region feature information for the feature region includes:

the feature region is normalized to a fixed size and then region feature information is extracted for the feature region.

Optionally, the candidate region is a rectangular region.

Optionally, extracting candidate feature information for the candidate region includes:

the candidate regions are normalized to a fixed size and then candidate feature information is extracted for the candidate regions.

Optionally, the method further comprises:

and transferring the second image to other albums except the target object album.

Optionally, the method further comprises:

and carrying out selection or deletion processing on the second image.

According to a second aspect of the embodiments of the present disclosure, there is provided an image processing apparatus, the apparatus including:

the system comprises an acquisition module, a display module and a display module, wherein the acquisition module is configured to acquire a first image from a preset target object photo album; images in the target object photo album all comprise first characteristics;

a selection module configured to select a feature region containing a second feature in the first image based on the first feature;

an extraction module configured to extract region feature information for the feature region;

a searching module configured to search for a second image in images other than the target object album based on the region feature information; the second image includes features similar to the second features.

Optionally, the apparatus further comprises:

a moving module configured to move the second image into the target object album.

Optionally, the search module includes:

a first judging submodule configured to judge whether an image other than the target object album contains a feature similar to the second feature in a sliding window manner based on the region feature information; and

a second judgment sub-module configured to determine that an image other than the target object album is the second image when the first judgment sub-module judges that the image contains a feature similar to the second feature.

Optionally, the search module includes:

a first selection sub-module configured to select a candidate area on an image other than the target object album;

a first extraction sub-module configured to extract candidate feature information for the candidate region;

a third judging submodule configured to judge whether the similarity between the candidate feature information and the region feature information is greater than a threshold value; and

a fourth judgment sub-module configured to determine that the image having the candidate region contains a feature similar to the second feature when the third judgment sub-module judges that the degree of similarity between the candidate feature information and the region feature information is greater than the threshold.

Optionally, the first feature is a face of the target object.

Optionally, the apparatus further comprises:

an identification module configured to identify age information and/or gender information of a plurality of persons for faces of the plurality of persons when the faces of the plurality of persons are included in the first image; and

a first judging module configured to determine a face of the target object according to the age information and/or gender information of the plurality of persons identified by the identifying module.

Optionally, the second characteristic is clothing of the target object.

Optionally, the apparatus further comprises:

the grouping module is configured to group all the images inside and outside the target object album according to the attributes of the images; and

the second judging module is configured to judge whether images within the target object album exist in each group of grouped images;

wherein the acquisition module is configured to acquire the first image from images existing in both the group of images and the target object album when the second determination module determines that an image within the target object album exists in the group of images.

Optionally, the second image is present in the set of images.

Optionally, the feature region is a rectangular region.

Optionally, the extraction module is configured to:

Optionally, the candidate region is a rectangular region.

Optionally, the first extraction submodule is configured to:

Optionally, the apparatus further comprises:

and the moving module is configured to move and store the second image into other albums except the target object album.

Optionally, the apparatus further comprises:

and the processing module is configured to select or delete the second image.

According to a third aspect of the embodiments of the present disclosure, there is provided an image processing apparatus including:

a processor;

a memory for storing processor-executable instructions;

wherein the processor is configured to:

extracting region feature information for the feature region;

The technical scheme provided by the embodiment of the disclosure can have the following beneficial effects:

the disclosed embodiments provide an image processing technique capable of selecting a feature region including a second feature based on a first feature and extracting region feature information when a first image is aggregated by a first feature in a target object album of a user, thereby being capable of searching for an image having a feature similar to the second feature outside the album based on the region feature information, and thus being capable of avoiding missing images of the target object to the greatest extent. Therefore, the image processing technology enables a user to automatically search the images outside the target object album which belong to the same target object as the images in the target object album without manually traversing all the images to search the images related to the target object album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the disclosure.

Drawings

The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and together with the description, serve to explain the principles of the disclosure.

FIG. 1 is a flow chart illustrating a method of image processing according to an exemplary embodiment of the present disclosure;

FIG. 2 is an effect diagram for illustrating an application scenario according to a specific embodiment of the present disclosure;

FIG. 3 is a flowchart illustrating one implementation of step S140 of FIG. 1, according to another exemplary embodiment of the present disclosure;

FIG. 4 is a flowchart illustrating one implementation of step S140 of FIG. 1, according to another exemplary embodiment of the present disclosure;

fig. 5 is a flow of an implementation of determining a face of a target object among a plurality of faces in an image processing method according to another exemplary embodiment of the present disclosure;

fig. 6 is an implementation flow of acquiring a first image in an image processing method according to another exemplary embodiment of the present disclosure;

FIG. 7 is a flow diagram illustrating a method of image processing according to another specific embodiment of the present disclosure;

FIG. 8 is a block diagram illustrating an image processing apparatus according to an exemplary embodiment of the present disclosure;

FIG. 9 is a block diagram illustrating the structure of the search module 840 of FIG. 8 according to an exemplary embodiment of the present disclosure;

fig. 10 is a block diagram illustrating a structure of a search module 840 in fig. 8 according to another exemplary embodiment of the present disclosure;

fig. 11 is a block diagram illustrating a portion of determining a face of a target object among a plurality of faces in an image processing apparatus according to another exemplary embodiment of the present disclosure;

fig. 12 is a block diagram illustrating a portion of an image processing apparatus acquiring a first image according to another exemplary embodiment of the present disclosure;

fig. 13 is a block diagram illustrating an image processing apparatus according to an exemplary embodiment of the present disclosure;

fig. 14 is a block diagram illustrating another image processing apparatus according to an exemplary embodiment of the present disclosure.

Detailed Description

Reference will now be made in detail to the exemplary embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the exemplary embodiments below are not intended to represent all implementations consistent with the present disclosure. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present disclosure, as detailed in the appended claims.

Fig. 1 is a flowchart illustrating an image processing method according to an exemplary embodiment of the present disclosure, the method including the following steps S110-S140:

in step S110, a first image is acquired from a preset target object album; the images in the target object album all contain the first feature.

In one embodiment, the first feature may be a face of the target object. In this case, the preset target object album may be formed by placing photos having the same person's face among the photos of the user in the same album by using face recognition and face clustering.

Fig. 5 is a flow of an implementation of determining a face of a target object among a plurality of faces in an image processing method according to another exemplary embodiment of the present disclosure.

As shown in fig. 5, in one embodiment, the image processing method according to another exemplary embodiment of the present disclosure may further include steps S510 and S520. In step S510, when faces of a plurality of persons are included in the first image, age information and/or gender information of the plurality of persons is identified for the faces of the plurality of persons. In step S520, a face of the target object is determined according to the age information and/or gender information of the plurality of identified persons. Through the flow shown in fig. 5, an image processing method according to another exemplary embodiment of the present disclosure can quickly and accurately determine a face of a target object, i.e., a first feature, in an image containing faces of a plurality of persons.

In step S120, a feature region including the second feature is selected in the first image based on the first feature.

In one embodiment, the second characteristic is clothing of the target object. In this case, a feature region of the clothing including the target object may be selected in the first image based on the first feature, for example, the face of the target object.

In one embodiment, the feature region includes a first feature and a second feature. In another embodiment, the feature region may not contain the first feature but contain the second feature. For example, when the first feature is a face of the target object and the second feature is clothing of the target object, the feature area contains clothing of the target object, since the target object album is set up based on the face of the target object, and images other than the target object album may not contain the face of the target object but may contain clothing of the target object. In other words, the feature region contains the second feature, so that in the method according to the embodiment of the present disclosure, an image containing a feature similar to the second feature can be searched for.

In step S130, region feature information is extracted for the feature region.

In one embodiment, the feature area is a rectangular area. In some cases, rectangular regions are advantageous for performing operations such as feature extraction. In other embodiments, other shapes of regions are possible, such as circular, elliptical, or polygonal feature regions other than rectangular. In other words, any shape of feature region is possible as long as feature information extraction is possible.

In one embodiment, step S130 includes: the feature region is normalized to a fixed size, and then region feature information is extracted for the feature region. Normalization makes the image resistant to attacks by geometric transformations and enables finding those invariants in the image. Therefore, the feature information of the feature region is extracted after normalization to a fixed size, so that searching for an image using such obtained feature information is more accurate.

In step S140, searching for a second image in images other than the target object album based on the region feature information; the second image contains features similar to the second features.

The above-described image processing method as shown in fig. 1 is further explained below with reference to an application scenario shown in fig. 2.

Fig. 2 is an effect diagram for explaining an application scenario according to a specific embodiment of the present disclosure. In FIG. 2, the target object album 200 contains

images

220 and 230 of the target object 210, and does not contain image 240. In addition, the target object album 200 shown in fig. 2 includes two

images

220 and 230 only as an example, and the target object album 200 may include one image and may include three or more images.

When the above-described image processing method shown in fig. 1 is applied to the application scenario shown in fig. 2, in step S110, a first image 220 is acquired from a preset target object album 200; the

images

220 and 230 in the target object album 200 each contain a first feature 221, namely the face of the target object 210. As is apparent from the above description of the above-described image processing method with reference to fig. 1, when images are aggregated with the faces of the target object 210 in the target object album 200, the images 240 in which the faces of the target object 210 are not photographed are not aggregated in the target object album 200.

In step S120, a feature region 225 containing a second feature 222, i.e., the clothing of the target object 210, is selected in the first image 220 based on the first feature 221. As is apparent from the above description of the above-described image processing method with reference to fig. 1, although the feature region 225 shown in fig. 2 includes both the first feature 221 and the second feature 222, the feature region 225 may include the second feature 222 without including the first feature 221. It is noted that although the

images

220 and 230 in fig. 2 are identical, this is merely an example and the

images

220 and 230 may have the same first feature 221, and need not be identical.

In step S130, region feature information is extracted for the feature region 225. In the embodiment shown in fig. 2, the feature region 225 is a rectangular region, but it is understood that the disclosure is not so limited.

In step S140, a second image 240 is searched for among images other than the target object album 200 based on the region feature information; the second image 240 contains a feature 242 that is similar to the second feature 222.

As can be seen from the above-described image processing method shown in fig. 1 explained with reference to the application scenario shown in fig. 2, when the target object album 200 of the user aggregates the first image by the first feature 221, the image processing method can select a feature region 225 including the second feature 222 based on the first feature 221 and extract region feature information, so that the image having the second feature 222 can be searched outside the target object album 200 based on the region feature information, and thus the missing of the image of the target object can be avoided to the maximum extent. Therefore, the image processing method enables a user to automatically search the images outside the target object photo album which belong to the same target object as the images in the target object photo album without manually traversing all the images to search the images related to the target object photo album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

It is to be noted that although the image processing method described above with reference to fig. 1 and 2 may have the first image 220 with the face of the target object as the first feature 221 and the clothing of the target object as the second feature 222, the present disclosure is not limited thereto. For example, the first feature may be the clothing 222 of the target object and the second feature is the face 221 of the target object. As another example, the first characteristic is the clothing 222 of the target subject, and the second characteristic is a hat (not shown in the figures) worn by the target subject. As another example, the first characteristic is a face of the target object, and the second characteristic is a bouquet or balloon (not shown) held by the target object. In other words, the first feature and the second feature may be any two non-identical features in the image.

Another implementation flow of step S140 in fig. 1 is described below with reference to fig. 3 and 4.

Fig. 3 is a flowchart illustrating an implementation of step S140 in fig. 1 according to another exemplary embodiment of the present disclosure. As shown in fig. 3, step S140 in fig. 1 may further include steps S1401 and S1402.

In step S1401, it is determined whether or not an image other than the target object album includes a feature similar to the second feature in the form of a sliding window based on the region feature information.

In step S1402, when it is determined that an image other than the target object album contains a feature similar to the second feature, it is determined that the image is the second image.

Steps S1401 and S1402 shown in fig. 3 are explained below with reference to the application scenario of fig. 2.

In step S1401, it is determined whether or not an image other than the target object album contains a feature similar to the second feature 222 in the form of a sliding window (for example, the sliding window 245 shown in fig. 2) based on the region feature information.

In step S1402, when it is determined that an image other than the target object album contains the feature 242 similar to the second feature 222, it is determined that the image 240 is the second image.

As can be seen from an implementation flow of step S140 shown in fig. 3 and described with reference to the application scenario shown in fig. 2, when the target object album 200 is searched for the image having the second feature 222 based on the regional feature information, whether the feature 242 similar to the second feature 222 exists in the image can be determined by using the sliding window, so that it is possible to avoid missing a possible similar feature 242 in one image, and thus, to avoid missing the image of the target object to the greatest extent. Therefore, the image processing technology enables a user to automatically search the images outside the target object album which belong to the same target object as the images in the target object album without manually traversing all the images to search the images related to the target object album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

Fig. 4 is a flowchart illustrating an implementation of step S140 in fig. 1 according to another exemplary embodiment of the present disclosure. As shown in fig. 4, step S140 in fig. 1 may further include steps S1403-S1406.

In step S1403, a candidate area is selected on an image other than the target album.

In step S1404, candidate feature information is extracted for the candidate region.

In one embodiment, the candidate region is a rectangular region. In some cases, rectangular regions are advantageous for performing operations such as feature extraction. In another embodiment, other shapes of the regions are possible, for example, candidates of polygons other than circles, ellipses or rectangles. In other words, any shape of candidate region is possible as long as feature information extraction is possible.

In one embodiment, step S1404 includes: the candidate regions are normalized to a fixed size and then candidate feature information is extracted for the candidate regions. Normalization makes the image resistant to attacks by geometric transformations and enables finding those invariants in the image. Therefore, candidate feature information of the candidate region is extracted after normalization to a fixed size, so that searching for an image using such obtained candidate feature information is more accurate.

In step S1405, it is determined whether the similarity between the candidate feature information and the region feature information is greater than a threshold. The threshold value may be determined based on a number of experiments and analyses, and is not particularly limited herein.

In step S1406, when the degree of similarity between the candidate feature information and the region feature information is greater than the threshold value, it is determined that the image having the candidate region contains a feature similar to the second feature.

The steps S1403 to S1406 shown in fig. 4 are explained below with reference to the application scenario of fig. 2.

In step S1403, a candidate area 245 is selected on the image 240 other than the target object album 200.

In step S1404, candidate feature information is extracted for the candidate region 245.

In step S1405, it is determined whether the similarity between the candidate feature information and the region feature information is greater than a threshold.

In step S1406, when the similarity between the candidate feature information and the region feature information is greater than the threshold, it is determined that the image 240 having the candidate region contains the feature 242 similar to the second feature 222.

As can be seen from an implementation flow of step S140 shown in fig. 4 and described with reference to the application scenario shown in fig. 2, when searching for a feature similar to the second feature 222 outside the target object album 200 based on the region feature information, the candidate feature region 245 is selected on the image outside the target object album and the candidate feature information is extracted for the candidate region 245, so that the similarity between the feature region 225 and the candidate feature region 245 can be accurately determined by comparing the similarities between the region feature information and the candidate feature information, and further, whether or not the image outside the target object album has the feature 242 similar to the second feature 222 can be accurately determined. Thus, it is possible to accurately judge whether one image is the second image. Therefore, the image processing technology enables a user to automatically search the images outside the target object album which belong to the same target object as the images in the target object album without manually traversing all the images to search the images related to the target object album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

One implementation flow of acquiring a first image in an image processing method according to another exemplary embodiment of the present disclosure is described below with reference to fig. 6.

Fig. 6 is an implementation flow of acquiring a first image in an image processing method according to another exemplary embodiment of the present disclosure.

As shown in fig. 6, the image processing method according to another exemplary embodiment of the present disclosure further includes steps S610 to S630.

In step S610, all images inside and outside the target object album are grouped according to the attribute of the image.

In step S620, it is determined whether or not an image within the target object album exists in each group of images obtained by grouping.

In step S630, when it is determined that an image within the target object album exists in a group of images, a first image is acquired from images existing both in the group of images and within the target object album.

In one embodiment, the second image is present in the set of images. In other words, the second image is present both in the set of images and outside the target object album.

In one embodiment, the attributes of the image include at least one of a photographing time of the image, an image size, an image resolution, an image bit size, and a file format of the image.

By grouping all the images inside and outside the target object album according to the attributes of the images, the image processing method of the embodiment can improve the probability of searching the second image, can narrow the range of the image to be searched, and can improve the speed of searching the second image.

For example, there are 40 images inside the target image album and 60 images outside the target image album for a total of 100 images. With the image processing method according to another exemplary embodiment of the present disclosure, in step S610, the 100 images are divided into, for example, 5 time periods, i.e., 5 groups, according to the capturing time period of the images, and the number of the images in each group may or may not be equal. In step S620, for example, a group of 20 images is selected, and it is determined whether or not an image within the target object album exists in the group of 20 images. In step S630, when it is determined that there are 5 images within the target object album in the set of 20 images, the first image is acquired from the 5 images. Since the user is likely to take more than one image over a period of time, the second characteristic of the user (e.g., clothing) will generally not change over a period of time. Thus, when there are 5 first images in the set of 20 images, it is illustrated that there is a high probability that there is an image capturing a second feature (e.g., clothing) of the target object in the remaining 15 images in the set of 20 images. Obviously, the probability of searching for the second image in the remaining 15 images in the group is higher than the probability of searching for the second image in 60 images other than the entire target object album, and a decrease in the amount of search indicates a faster search speed.

For another example, if all of the above-mentioned 20 images in a group are images in the target object album, the image processing method according to another exemplary embodiment of the present disclosure may not select the first image in the 20 images because there are no images other than the target object album in the same time period of the group of 20 images, and thus the probability that the second image is searched in all 60 images other than the target object album is low. Obviously, not acquiring the first image within the set of 20 images, but acquiring the first image from images existing in both the set of images and the target object album, can both improve the probability of searching for the second image and reduce the range of the image to be searched, and can also improve the speed of searching for the second image.

Although the above description has been made only for grouping the images according to the capturing time of the images, the images may be grouped according to at least one of the image size, the image resolution, the image bit size, and the file format of the images. By grouping all the images inside and outside the target object album according to the attributes of the images, the image processing method of the embodiment can improve the probability of searching for the second image, can narrow the range of the images to be searched, and can improve the speed of searching for the second image.

In an embodiment of the present disclosure, the image processing method described above may further include the steps of: and moving the second image into the target object album. The searched second image is moved into the target object photo album, so that the accuracy and convenience of image arrangement can be improved, and great convenience is brought to the use of a user.

In an embodiment of the present disclosure, the image processing method described above may further include the steps of: and transferring the second image to other albums except the target object album. The searched second image is transferred to other albums except the target object album, so that the image sorting accuracy and convenience can be improved, and great convenience is brought to the use of a user.

In an embodiment of the present disclosure, the image processing method described above may further include the steps of: and performing selection or deletion processing on the second image. The searched second image is selected or deleted, so that the accuracy and convenience of image sorting can be improved, and great convenience is brought to the use of a user.

FIG. 7 is a flow chart illustrating an image processing method according to another specific embodiment of the present disclosure.

In the image processing method of this embodiment, the entire image processing flow S710 to S765 is performed for a specific application scenario, i.e., a finishing baby album (i.e., an album that holds images of one specific baby/child).

As shown in fig. 7, in step S710, a baby album is generated by face recognition and face clustering. That is, most of the baby images will be gathered together by the baby album first. For example, firstly, face recognition and face clustering methods are adopted, and images of faces of the same baby in the user's image are placed in the same album. Images containing faces of different people may be placed in different albums. Specifically, there are two ways of creating a baby album by automatically identifying a baby and manually creating a baby album by a user.

In step S715, a baby in the image is determined by age identification and/or gender identification. After clustering is finished, a gender and/or age identification method can be adopted, the face of each photo album is identified, and if the age is smaller than a certain threshold value, the photo album is determined to be a baby.

In step S720, the images are grouped at regular time intervals and each grouped image is analyzed. When taking a picture, the baby is not usually in a matching condition, such as lying on the back, taking a shadow, turning over, and the like. This adds difficulty to face detection, so overall, the photo recall rate for babies is lower than for average adults. But if it is a parent of the baby, the baby's shadow picture is also generally preferred. Therefore, the method based on image searching can be adopted to increase the recall rate of the baby. This is because the clothes of babies and the like are generally not changed at the same time.

In step S725, it is determined whether or not a baby image exists in the packet.

In step S730, when it is determined that there is a baby image in the group, the face area of the baby is enlarged, and a rectangular area including the face and the clothes of the baby is determined. When it is judged that the baby image does not exist in the group, the image processing method is ended.

In step S735, the rectangular region is normalized to a fixed size, and then region feature information is extracted for the region. For example, the region feature information may be feature information such as a convolution feature and a color texture feature trained by a Convolution Neural Network (CNN).

In step S740, the other image selection candidate frames in the group other than the baby album are traversed using the sliding window. For example, the search determines that the photos of the baby are not contained in the other photos in the group (determined in the baby photo album), and then selects candidate frames with different sizes for traversing from top to bottom and from left to right for each picture in the form of a sliding window.

In step S745, each candidate frame is selected, normalized to a fixed size, and candidate feature information is extracted for the region.

In step S750, the candidate feature information and the region feature information are compared, and the similarity between the candidate feature information and the region feature information is determined. That is, the similarity of the features in the rectangular area including the baby face, clothes, and the features in the candidate frame is compared.

In step S755, it is determined whether the similarity between the candidate feature information and the region feature information is greater than a threshold.

In step S760, when it is determined that the similarity between the candidate feature information and the region feature information is greater than the threshold, which indicates that the features in the rectangular region including the face and the clothes of the baby are similar to the features in the candidate frame, the candidate frame is the searched baby region, and the image is moved into a baby album. One image theoretically has only one target baby, and after the target baby is found, the image searching is finished. When it is determined that the similarity between the candidate feature information and the region feature information is not greater than the threshold, the search of this image is also ended.

In step S765, it is determined whether all images have been traversed. When it is determined that all of the images have been traversed, a method of image processing according to another specific embodiment of the present disclosure, shown in FIG. 7, ends. When it is determined that all the images have not been traversed, return is made to step S740. At this time, this baby in each group is searched in the method of the above-described steps S740 to S765. And then adding the searched image into the baby photo album, thereby improving the recall rate of the baby photo album.

In another image processing method according to another embodiment of the present disclosure, a majority of babies are gathered together through a baby album, then a rectangular area of a frame of each baby in the baby album is obtained, then features are extracted, and other photos, such as a back shadow and the like, in a similar time period are searched, so that a recall rate of the baby album is increased.

When the baby photo album is created, the face can be detected firstly, then the classification model is adopted to judge whether the face is a baby or not, and if the face is a baby, the face is added into the baby photo album. Because the baby is disorderly or uncoordinated frequently, or the face of the baby is shielded, and the like, the face of the baby can not be detected basically or can be missed. For example, a total of 100 faces, 60 regular faces of babies, 40 other occluded faces or the same baby without face has a normal recall rate of 60%. When a user takes a picture, the user generally takes more than one picture in a time period, and after the face of a baby is normally detected, the clothes area of the baby is cut out. Based on its clothing area, 40 pictures of undetected babies were searched. Therefore, the recall rate of the baby photo album is improved.

In one embodiment, if it is known what clothing a baby is currently wearing, similar faces of a baby that may exist can be searched from other images in which a baby is not detected. The method for searching similar images is realized by firstly aggregating human face books, identifying a baby according to age, then intercepting information of a clothes area of the baby, and then identifying whether the baby possibly exists in other pictures by adopting a sliding window and image searching (which can be CNN or Support Vector Machine (SVM)) mode.

In one embodiment, a baby is identified by age, information of the clothes area of the baby is intercepted, and whether the baby possibly exists is identified from other pictures by adopting a sliding window + image search (which can be CNN or SVM and the like). The method is equivalent to that the clothes information is a condition, and then an object detection method is introduced, wherein the object detection of the baby is dynamically changed along with the baby object.

Fig. 8 is a block diagram illustrating an image processing apparatus according to an exemplary embodiment of the present disclosure, which includes the following acquisition module 810, selection module 820, extraction module 830, and search module 840.

The obtaining module 810 is configured to obtain a first image from a preset target object album; the images in the target object album all contain the first feature.

Fig. 11 is a block diagram illustrating a portion for determining a face of a target object among a plurality of faces in an image processing apparatus according to another exemplary embodiment of the present disclosure.

As shown in fig. 11, in one embodiment, the image processing apparatus according to another exemplary embodiment of the present disclosure may further include an identification module 1110 and a first judgment module 1120. The identifying module 1110 is configured to identify age information and/or gender information of a plurality of persons for faces of the plurality of persons when the faces of the plurality of persons are included in the first image. The first judging module 1120 is configured to determine the face of the target object according to the age information and/or the gender information of the plurality of persons identified by the identifying module 1110. Through the block diagram shown in fig. 11, an image processing apparatus according to another exemplary embodiment of the present disclosure can quickly and accurately determine a face of a target object, i.e., a first feature, in an image containing faces of a plurality of persons.

The selection module 820 is configured to select a feature region containing a second feature in the first image based on the first feature.

The extraction module 830 is configured to extract region feature information for the feature region.

In one embodiment, the extraction module 830 is configured to normalize the feature region to a fixed size and then extract region feature information for the feature region. Normalization makes the image resistant to attacks by geometric transformations and enables finding those invariants in the image. Therefore, the feature information of the feature region is extracted after normalization to a fixed size, so that searching for an image using such obtained feature information is more accurate.

The search module 840 is configured to search for a second image in images other than the target object album based on the region feature information; the second image contains features similar to the second features.

The image processing apparatus as described above with reference to fig. 8 is further described below with reference to an application scenario as shown in fig. 2.

images

When the above-described image processing apparatus shown in fig. 8 is applied to the application scenario shown in fig. 2, the obtaining module 810 is configured to obtain the first image 220 from the preset target object album 200; the

images

220 and 230 in the target object album 200 each contain a first feature 221, namely the face of the target object 210. As is apparent from the above description of the image processing apparatus with reference to fig. 8, when images are aggregated with the faces of the target object 210 in the target object album 200, the images 240 in which the faces of the target object 210 are not photographed are not aggregated in the target object album 200.

The selection module 820 is configured to select the feature region 225 in the first image 220 containing the second feature 222, i.e. the clothing of the target object 210, based on the first feature 221. As is apparent from the above description of the image processing apparatus with reference to fig. 8, although the feature region 225 shown in fig. 2 includes both the first feature 221 and the second feature 222, the feature region 225 may include the second feature 222 without including the first feature 221. It is noted that although the

images

220 and 230 in fig. 2 are identical, this is merely an example and the

images

220 and 230 may have the same first feature 221, and need not be identical.

The extraction module 830 is configured to extract region feature information for the feature region 225. In the embodiment shown in fig. 2, the feature region 225 is a rectangular region, but it is understood that the disclosure is not so limited.

The search module 840 is configured to search for the second image 240 in images other than the target object album 200 based on the region feature information; the second image 240 contains a feature 242 that is similar to the second feature 222.

As is apparent from the above-described image processing apparatus shown in fig. 8 explained with reference to the application scenario shown in fig. 2, when the target object album 200 of the user aggregates first images by the first features 221, the image processing apparatus can select a feature region 225 including the second features 222 based on the first features 221 and extract region feature information, and can search for an image having the features 242 similar to the second features 222 outside the target object album 200 based on the region feature information, so that omission of the image of the target object can be avoided to the greatest extent. Therefore, the image processing device enables a user to automatically search the images outside the target object album which belong to the same target object as the images in the target object album without manually traversing all the images to search the images related to the target object album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

It is to be noted that although the image processing apparatus described above with reference to fig. 8 and 2 may have the first image 220 with the face of the target object as the first feature 221 and the clothing of the target object as the second feature 222, the present disclosure is not limited thereto. For example, the first feature may be the clothing 222 of the target object and the second feature is the face 221 of the target object. As another example, the first characteristic is the clothing 222 of the target subject, and the second characteristic is a hat (not shown in the figures) worn by the target subject. As another example, the first characteristic is a face of the target object, and the second characteristic is a bouquet or balloon (not shown) held by the target object. In other words, the first feature and the second feature may be any two non-identical features in the image.

The following describes other structures of the search module 840 in fig. 8 with reference to fig. 9 and 10.

Fig. 9 is a block diagram illustrating a structure of the search module 840 of fig. 8 according to an exemplary embodiment of the present disclosure. As shown in fig. 9, the search module 840 of fig. 8 may further include a first judgment sub-module 8401 and a second judgment sub-module 8402.

The first judgment sub-module 8401 is configured to judge whether or not an image other than the target object album contains a feature similar to the second feature in a sliding window manner based on the region feature information.

The second judging submodule 8402 is configured to determine that an image other than the target object album is a second image when the first judging submodule 8401 judges that the image contains a feature similar to the second feature.

The first and

second determination sub-modules

8401 and 8402 shown in fig. 9 are explained below with reference to an application scenario of fig. 2.

The first judgment sub-module 8401 is configured to judge whether or not an image other than the target object album contains a feature similar to the second feature 222 in the form of a sliding window (for example, the sliding window 245 shown in fig. 2) based on the region feature information.

The second judging sub-module 8402 is configured to determine that the image 240 is the second image when the first judging sub-module 8401 judges that an image other than the target object album contains the feature 242 similar to the second feature 222.

As can be seen from the structure of the search module 840 shown in fig. 9 explained with reference to the application scenario shown in fig. 2, when searching for an image having a feature 242 similar to the second feature 222 outside the target object album 200 based on the regional feature information, it is possible to determine whether there is a feature 242 similar to the second feature 222 in the image by using a sliding window, so that it is possible to avoid missing a possible similar feature 242 in one image, and thus, to avoid missing the image of the target object to the greatest extent. Therefore, the image processing technology enables a user to automatically search the images outside the target object album which belong to the same target object as the images in the target object album without manually traversing all the images to search the images related to the target object album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

Fig. 10 is a block diagram illustrating a structure of the search module 840 of fig. 8 according to another exemplary embodiment of the present disclosure. As shown in fig. 10, the search module 840 of fig. 8 may further include a first selection sub-module 8403, a first extraction sub-module 8404, a third judgment sub-module 8405, and a fourth judgment sub-module 8406.

The first selection submodule 8403 is configured to select a candidate area on an image other than the target object album.

The first extraction sub-module 8404 is configured to extract candidate feature information for the candidate region.

In one embodiment, the first extraction sub-module 8404 is configured to normalize the candidate region to a fixed size and then extract candidate feature information for the candidate region. Normalization makes the image resistant to attacks by geometric transformations and enables finding those invariants in the image. Therefore, candidate feature information of the candidate region is extracted after normalization to a fixed size, so that searching for an image using such obtained candidate feature information is more accurate.

The third judgment sub-module 8405 is configured to judge whether the similarity of the candidate feature information and the area feature information is greater than a threshold value. The threshold value may be determined based on a number of experiments and analyses, and is not particularly limited herein.

The fourth judging submodule 8406 is configured to determine that the image having the candidate region contains a feature similar to the second feature when the third judging submodule 8405 judges that the degree of similarity of the candidate feature information and the region feature information is larger than the threshold value.

The first selection sub-module 8403, the first extraction sub-module 8404, the third judgment sub-module 8405, and the fourth judgment sub-module 8406 shown in fig. 10 are explained below with reference to the application scenario of fig. 2.

The first selection sub-module 8403 is configured to select a candidate area 245 on the image 240 outside the target object album 200.

The first extraction sub-module 8404 is configured to extract candidate feature information for the candidate region 245.

The third judgment sub-module 8405 is configured to judge whether the similarity of the candidate feature information and the area feature information is greater than a threshold value.

The fourth judging submodule 8406 is configured to determine that the image 240 having the candidate region contains the feature 242 similar to the second feature 222 when the third judging submodule 8405 judges that the degree of similarity of the candidate feature information and the region feature information is larger than the threshold value.

As is clear from the structure of the search module 840 shown in fig. 10 described with reference to the application scenario shown in fig. 2, when searching for an image having a feature 242 similar to the second feature 222 outside the target object album 200 based on the regional feature information, by selecting the candidate feature region 245 on the image outside the target object album and extracting the candidate feature information for the candidate region 245, it is possible to accurately judge the similarity between the feature region 225 and the candidate feature region 245 by comparing the similarities between the regional feature information and the candidate feature information, and thus it is possible to accurately judge whether or not the image outside the target object album has the feature 242 similar to the second feature 222. Thus, it is possible to accurately judge whether one image is the second image. Therefore, the image processing technology enables a user to automatically search the images outside the target object album which belong to the same target object as the images in the target object album without manually traversing all the images to search the images related to the target object album, thereby improving the accuracy and convenience of image arrangement and bringing great convenience to the user.

A portion of acquiring a first image in an image processing apparatus according to another exemplary embodiment of the present disclosure is described below with reference to fig. 12.

Fig. 12 is a block diagram illustrating a portion of an image processing apparatus acquiring a first image according to another exemplary embodiment of the present disclosure.

As shown in fig. 12, the image processing apparatus according to another exemplary embodiment of the present disclosure further includes a grouping module 1210, a second determination module 1220, and an acquisition module 810.

The grouping module 1210 is configured to group all images inside and outside the target object album by their attributes.

The second determination module 1220 is configured to determine whether an image within the target object album exists in each group of images obtained by grouping.

The obtaining module 810 is configured to obtain the first image from images existing in both the set of images and the target object album when the second judging module 1220 judges that an image within the target object album exists in the set of images.

By grouping all the images inside and outside the target object album according to the attributes of the images, the image processing apparatus of the embodiment can not only increase the probability of searching for the second image, but also reduce the range of the images to be searched, and can also increase the speed of searching for the second image.

For example, there are 40 images inside the target image album and 60 images outside the target image album for a total of 100 images. With the image processing apparatus according to another exemplary embodiment of the present disclosure, the grouping module 1210 is configured to divide the 100 images into, for example, 5 time periods, i.e., 5 groups, according to the capturing time period of the images, and the number of the images in each group may or may not be equal. The second determination module 1220 is configured to select a group of 20 images, for example, and determine whether an image within the target object album exists in the group of 20 images. The obtaining module 810 is configured to obtain the first image from the 5 images when the second judging module 1220 judges that there are 5 images within the target object album in the set of 20 images. Since the user is likely to take more than one image over a period of time, the second characteristic of the user (e.g., clothing) will generally not change over a period of time. Thus, when there are 5 first images in the set of 20 images, it is illustrated that there is a high probability that there is an image capturing a second feature (e.g., clothing) of the target object in the remaining 15 images in the set of 20 images. Obviously, the probability of searching for the second image in the remaining 15 images in the group is higher than the probability of searching for the second image in 60 images other than the entire target object album, and a decrease in the amount of search indicates a faster search speed.

For another example, if all of the above-described set of 20 images are images within the target object album, the image processing apparatus according to another exemplary embodiment of the present disclosure may not select the first image within the 20 images because there are no images outside the target object album within the same time period of the set of 20 images, and thus the probability that the second image is searched for in all 60 images outside the target object album is low. Obviously, not acquiring the first image within the set of 20 images, but acquiring the first image from images existing in both the set of images and the target object album, can both improve the probability of searching for the second image and reduce the range of the image to be searched, and can also improve the speed of searching for the second image.

Although the above describes grouping the images according to the capturing time of the images, the images may be grouped according to at least one of the image size, the image resolution, the image bit size, and the file format of the images. By grouping all the images inside and outside the target object album according to the above-described attributes of the images, the image processing apparatus of the present embodiment can increase the probability of searching for the second image, and can also narrow the range of the images to be searched, and can also increase the speed of searching for the second image.

In an embodiment of the present disclosure, the image processing apparatus described above may further include a moving module (not shown in the figure) configured to move the second image into the target object album. The searched second image is moved into the target object photo album, so that the accuracy and convenience of image arrangement can be improved, and great convenience is brought to the use of a user.

In an embodiment of the present disclosure, the image processing apparatus described above may further include a moving module (not shown in the figure) configured to move the second image to another album other than the target object album. The searched second image is transferred to other albums except the target object album, so that the image sorting accuracy and convenience can be improved, and great convenience is brought to the use of a user.

In an embodiment of the present disclosure, the image processing apparatus described above may further include a processing module (not shown) configured to perform a selecting or deleting process on the second image. The searched second image is selected or deleted, so that the accuracy and convenience of image sorting can be improved, and great convenience is brought to the use of a user.

An embodiment of the present disclosure provides an image processing apparatus, including:

a processor;

a memory for storing processor-executable instructions;

wherein the processor is configured to:

acquiring a first image from a preset target object photo album; images in the target object album all contain first characteristics;

extracting region feature information for the feature region;

searching for a second image in images other than the target object album based on the region feature information; the second image contains features similar to the second features.

Fig. 13 is a block diagram illustrating an image processing apparatus 1300 according to an exemplary embodiment of the present disclosure. For example, the apparatus 1300 may be a client, which may be an application program, or a mobile device, such as a mobile phone, a computer, a digital broadcast terminal, a messaging device, a game console, a tablet device, a medical device, an exercise device, a personal digital assistant, and the like.

Referring to fig. 13, the apparatus 1300 may include one or more of the following components: a processing component 1302, a memory 1304, a power component 1306, a multimedia component 1308, an audio component 1310, an input/output (I/O) interface 1312, a sensor component 1314, and a communication component 1316.

The processing component 1302 generally controls overall operation of the device 1300, such as operations associated with display, telephone calls, data communications, camera operations, and recording operations. The processing component 1302 may include one or more processors 1320 to execute instructions to perform all or part of the steps of the methods described above. Further, the processing component 1302 can include one or more modules that facilitate interaction between the processing component 1302 and other components. For example, the processing component 1302 may include a multimedia module to facilitate interaction between the multimedia component 1308 and the processing component 1302.

The memory 1304 is configured to store various types of data to support operations at the apparatus 1300. Examples of such data include instructions for any application or method operating on device 1300, contact data, phonebook data, messages, pictures, videos, and so forth. The memory 1304 may be implemented by any type or combination of volatile or non-volatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disks.

Power supply component 1306 provides power to the various components of device 1300. Power components 1306 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power for device 1300.

The multimedia component 1308 includes a screen between the device 1300 and the user that provides an output interface. In some embodiments, the screen may include a Liquid Crystal Display (LCD) and a Touch Panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive an input signal from a user. The touch panel includes one or more touch sensors to sense touch, slide, and gestures on the touch panel. The touch sensor may not only sense the boundary of a touch or slide action, but also detect the duration and pressure associated with the touch or slide operation. In some embodiments, the multimedia component 1308 includes a front facing camera and/or a rear facing camera. The front camera and/or the rear camera may receive external multimedia data when the apparatus 1300 is in an operation mode, such as a photographing mode or a video mode. Each front camera and rear camera may be a fixed optical lens system or have a focal length and optical zoom capability.

The audio component 1310 is configured to output and/or input audio signals. For example, the audio component 1310 includes a Microphone (MIC) configured to receive external audio signals when the apparatus 1300 is in an operating mode, such as a call mode, a recording mode, and a voice recognition mode. The received audio signals may further be stored in the memory 1304 or transmitted via the communication component 1316. In some embodiments, the audio component 1310 also includes a speaker for outputting audio signals.

The I/O interface 1312 provides an interface between the processing component 1302 and peripheral interface modules, which may be keyboards, click wheels, buttons, etc. These buttons may include, but are not limited to: a home button, a volume button, a start button, and a lock button.

The sensor assembly 1314 includes one or more sensors for providing various aspects of state assessment for the device 1300. For example, the sensor assembly 1314 may detect the open/closed state of the device 1300, the relative positioning of components, such as a display and keypad of the device 1300, the sensor assembly 1314 may also detect a change in the position of the device 1300 or a component of the device 1300, the presence or absence of user contact with the device 1300, orientation or acceleration/deceleration of the device 1300, and a change in the temperature of the device 1300. The sensor assembly 1314 may include a proximity sensor configured to detect the presence of a nearby object without any physical contact. The sensor assembly 1314 may also include a light sensor, such as a CMOS or CCD image sensor, for use in imaging applications. In some embodiments, the sensor assembly 1314 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

The communication component 1316 is configured to facilitate communications between the apparatus 1300 and other devices in a wired or wireless manner. The apparatus 1300 may access a wireless network based on a communication standard, such as WiFi, 2G or 3G, or a combination thereof. In an exemplary embodiment, the communication component 1316 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communications component 1316 also includes a Near Field Communications (NFC) module to facilitate short-range communications. For example, the NFC module may be implemented based on Radio Frequency Identification (RFID) technology, infrared data association (IrDA) technology, Ultra Wideband (UWB) technology, Bluetooth (BT) technology, and other technologies.

In an exemplary embodiment, the apparatus 1300 may be implemented by one or more Application Specific Integrated Circuits (ASICs), Digital Signal Processors (DSPs), Digital Signal Processing Devices (DSPDs), Programmable Logic Devices (PLDs), Field Programmable Gate Arrays (FPGAs), controllers, micro-controllers, microprocessors or other electronic components for performing the above-described methods.

In an exemplary embodiment, a non-transitory computer readable storage medium comprising instructions, such as the memory 1304 comprising instructions, executable by the processor 1320 of the apparatus 1300 to perform the method described above is also provided. For example, the non-transitory computer readable storage medium may be a ROM, a Random Access Memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like.

A non-transitory computer readable storage medium having instructions therein, which when executed by a processor of a mobile terminal, enable the mobile terminal to perform an image processing method, the method comprising:

extracting region feature information for the feature region;

Fig. 14 is a block diagram illustrating another image processing apparatus according to an exemplary embodiment of the present disclosure. For example, the apparatus 1400 may be provided as a server. Referring to fig. 14, the apparatus 1400 includes a processing component 1422 that further includes one or more processors and memory resources, represented by memory 1432, for storing instructions, such as applications, that are executable by the processing component 1422. The application programs stored in memory 1432 may include one or more modules each corresponding to a set of instructions. Further, the processing component 1422 is configured to execute instructions to perform the above-described methods.

The device 1400 may also include a power component 1426 configured to perform power management of the device 1400, a wired or wireless network interface 1450 configured to connect the device 1400 to a network, and an input output (I/O) interface 1458. The apparatus 1400 may operate based on an operating system stored in the memory 1432, such as Windows Server, Mac OS XTM, UnixTM, LinuxTM, FreeBSDTM, or the like.

A non-transitory computer readable storage medium in which instructions, when executed by a processor of a server, enable the server to perform a method of image processing, the method comprising:

extracting region feature information for the feature region;

Other embodiments of the disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the disclosure disclosed herein. This application is intended to cover any variations, uses, or adaptations of the disclosure following, in general, the principles of the disclosure and including such departures from the present disclosure as come within known or customary practice within the art to which the disclosure pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the disclosure being indicated by the following claims.

It will be understood that the present disclosure is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. An image processing method, comprising:

selecting a feature region containing a second feature in the first image based on the first feature, the feature region including the second feature and not including the first feature;

extracting region feature information for the feature region;

2. The method of claim 1, further comprising:

and moving the second image into the target object album.

3. The method of claim 1, wherein searching for a second image in an image other than the target object album based on the region feature information comprises:

4. The method of claim 1, wherein searching for a second image in an image other than the target object album based on the region feature information comprises:

selecting a candidate area on an image other than the target object album;

extracting candidate feature information for the candidate region;

5. The method of claim 1, wherein the first feature is a face of a target object.

6. The method of claim 5, further comprising:

7. The method of claim 1, wherein the second characteristic is clothing of the target object.

8. The method of claim 1, further comprising:

9. The method of claim 8, wherein the second image is present in the set of images.

10. The method of claim 8, wherein the attributes of the image comprise at least one of a photographing time of the image, an image size, an image resolution, an image bit size, and a file format of the image.

11. The method of claim 1, wherein the feature region contains the first feature and the second feature.

12. The method of claim 1, wherein the feature region is a rectangular region.

13. The method of claim 1, wherein extracting region feature information for the feature region comprises:

14. The method of claim 4, wherein the candidate region is a rectangular region.

15. The method of claim 4, wherein extracting candidate feature information for the candidate region comprises:

16. The method of claim 1, further comprising:

17. The method of claim 1, further comprising:

and carrying out selection or deletion processing on the second image.

18. An image processing apparatus characterized by comprising:

a selection module configured to select, based on the first feature, a feature region containing a second feature in the first image, the feature region including the second feature and not including the first feature;

19. The apparatus of claim 18, further comprising:

20. The apparatus of claim 18, wherein the search module comprises:

21. The apparatus of claim 18, wherein the search module comprises:

22. The apparatus of claim 18, wherein the first feature is a face of the target object.

23. The apparatus of claim 22, further comprising:

24. The apparatus of claim 18, wherein the second characteristic is clothing of the target object.

25. The apparatus of claim 18, further comprising:

26. The apparatus of claim 25, wherein the second image is present in the set of images.

27. The apparatus of claim 25, wherein the attribute of the image comprises at least one of a photographing time of the image, an image size, an image resolution, an image bit size, and a file format of the image.

28. The apparatus of claim 18, wherein the feature region contains the first feature and the second feature.

29. The apparatus of claim 18, wherein the feature region is a rectangular region.

30. The apparatus of claim 18, wherein the extraction module is configured to:

31. The apparatus of claim 21, wherein the candidate region is a rectangular region.

32. The apparatus of claim 21, wherein the first extraction sub-module is configured to:

33. The apparatus of claim 18, further comprising:

34. The apparatus of claim 18, further comprising:

and the processing module is configured to select or delete the second image.

35. An image processing apparatus characterized by comprising:

a processor;

a memory for storing processor-executable instructions;

wherein the processor is configured to:

extracting region feature information for the feature region;

36. A computer-readable storage medium having stored thereon computer instructions, which when executed by a processor, perform the steps of the method of any one of claims 1 to 18.