WO2022057773A1

WO2022057773A1 - Image storage method and apparatus, computer device and storage medium

Info

Publication number: WO2022057773A1
Application number: PCT/CN2021/118045
Authority: WO
Inventors: 雷霆; 苏炜
Original assignee: 华为技术有限公司
Priority date: 2020-09-15
Filing date: 2021-09-13
Publication date: 2022-03-24
Also published as: CN114268730A

Abstract

An image storage method and apparatus, a computer device and a storage medium, relating to the technical field of data storage. The method comprises: acquiring a first image, selecting and storing at least one image from among the first image and a stored image set, and deleting unselected images. The stored image set contains one or more images. The images in the image set and the first image have the same photographed subject, the same photographed place, and a photographing time interval less than a preset threshold value. In the present application, not all of the images of the same photographed subject that are consecutively photographed at the same place within a short time are stored, only some of the images are selected and stored, such that a waste of storage space can be effectively reduced.

Description

Image storage method, apparatus, computer equipment and storage medium

This application claims the priority of the Chinese patent application with application number 202010968900.0 filed on September 15, 2020 and the title of the invention is "image storage method, device, computer equipment and storage medium", the entire contents of which are incorporated herein by reference Applying.

technical field

The present application relates to the technical field of data storage, and in particular, to a method, apparatus, computer device and storage medium for image storage.

Background technique

With the continuous development of video surveillance, view applications such as face retrieval and face tracking have been derived from video surveillance. The captured image can usually be processed as follows before implementing the view application.

The camera sends the real-time captured images to the storage system, such as Huawei's Intelligent Video Surveillance (IVS), which parses the received images to obtain the face images in the images, and the corresponding images of each face. facial feature data and structured data. The face feature data may be a feature vector, and the structured data includes the shooting time, shooting location, quality score, age, gender, etc. corresponding to the face image. Then, the face image, the corresponding face feature data, and the face structured data are stored correspondingly for use by subsequent view applications.

In the process of realizing this application, the inventor found that the related art has at least the following problems:

If a subject stays in the same location for a long time, cameras deployed at that location will continue to capture a large number of images that include the subject. In this way, when the IVS analyzes these images, a large number of face images of the photographed object will be obtained. These face images are very similar to each other, but they will occupy a large amount of storage space and cause a waste of storage space.

SUMMARY OF THE INVENTION

Embodiments of the present application provide an image storage method, apparatus, computer device, and storage medium. The technical solution is as follows:

In a first aspect, a method for image storage is provided, the method comprising:

Acquire the first image, and perform a deduplication operation on the first image and the stored image set, specifically including:

Select at least one image from the first image and the stored image set to save, wherein the stored image set contains one or more images, and the images in the image set and the first image have: the same shooting object , the same shooting location, and the shooting time interval less than a preset threshold, delete the unselected images.

In the solutions provided by the embodiments of the present application, the image may be a face image, and for the convenience of description, the following image is a face image for description. The shooting location is the location where the image shooting device is located, and the preset threshold may be 5 minutes to 10 minutes. The face image can be obtained after the image capture device performs face recognition on the captured video, and then sent by the image capture device to the intelligent video surveillance platform (IVS, Intelligent Video Surveillance), or it can be sent by the IVS to the image capture device. The video is obtained by face recognition. In addition, the face images acquired by IVS are acquired in the order of shooting time. For example, the shooting time corresponding to the face image acquired later is later than the shooting time corresponding to the face image acquired earlier. After each acquisition of a face image, if it is determined that there is a face image with the same shooting object and shooting location as the currently acquired face image, and the shooting time interval is less than the preset threshold, the currently acquired face image will be stored. The face image is added to the image set corresponding to the face image satisfying the above conditions. Specifically, when judging whether the two face images correspond to the same shooting object, the similarity of the feature data of the two can be calculated, and if the similarity is greater than a preset threshold, it is considered that the two correspond to the same shooting object.

For the face images in the above-mentioned image set after adding processing, at least one face image can be selected to be saved, and the rest can be deleted. In this way, a deduplication operation can be performed on the face images of the same object that are continuously photographed at the same location in a short period of time, and it is not necessary to store all of them, which effectively saves storage resources.

In a possible implementation, the method further includes:

Acquiring a second image, and establishing an image set including the second image if there is no third image having the same shooting object, the same shooting location, and a shooting time interval smaller than a preset threshold as the second image.

In the solution provided by the embodiment of the present application, whenever a face image is acquired, it can be judged first whether there is the same shooting object, the same shooting location, and the shooting time interval smaller than the preset threshold as the face image. If there is no face image with the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold as the face image, a new image set will be established, and the obtained face image will be The face image is added to the newly created image collection.

In a possible implementation manner, the triggering conditions of the above-mentioned deduplication operation may include:

Within a preset period of time after the first image is acquired, a fourth face image with the same shooting object, the same shooting location, and a shooting time interval smaller than the preset threshold as the first face image is not acquired.

In the solution provided by the embodiment of the present application, for an image set, if the first face image (the latest added face image in the image set) is obtained within a preset time period, the first face image is not obtained. If the face image has the same shooting object, the same shooting location, and the face image whose shooting time interval is less than the preset threshold, the image collection can be de-duplicated. The preset duration may be equal to the aforementioned preset threshold. For example, the deduplication time interval may be set to 10 minutes, and the corresponding preset duration may also be set to 10 minutes, or the preset duration may be set to be slightly larger than the aforementioned preset threshold.

A fifth face image with the same shooting object as the first face image and a different shooting point is obtained.

In the solution provided by the embodiment of the present application, in this case, it is indicated that the photographing object corresponding to the first face image has gone to another place, and the image set can be deduplicated.

It is determined that the number of face images included in the image set is greater than the target number, or it is determined that the size of the storage space occupied by the face images included in the image set is greater than the target storage value.

In the solution provided by the embodiment of the present application, the number of face images included in each image set may be detected according to a preset detection period, and if the number is greater than the preset number, a deduplication operation may be performed on the image set. The target quantity can be set according to actual needs, for example, the preset quantity can be set to 10 to 20 sheets. Alternatively, the size of the storage space occupied by the face images in each image set may be detected according to a preset detection period, and if it is greater than the target storage value, the image set may be deduplicated. Among them, the target storage value can be set according to actual needs, for example, can be set to 300KB to 600KB.

In a possible implementation, the method further includes:

Modify the shooting time of the selected face image to the target shooting time period, wherein the starting time in the target shooting time period is the earliest time in the shooting time of the image in the image regroup and the first image, and the target shooting time The end time in the shooting time period is the latest time in the shooting time of the images in the image set and the first image, and the structured data and face feature data of the unselected images are deleted.

In the solution provided by the embodiment of the present application, in order to further save storage space, when deleting unselected face images, the structured data and face feature data corresponding to these face images may also be deleted. In addition, for retaining at least one face image in the removal set, its shooting time can be modified to the target shooting time period, wherein the starting time in the target shooting time period is the shooting time corresponding to the face image in the image set. The earliest time in the target shooting time period is the latest time in the shooting time corresponding to the face image in the image set. In this way, the saved face image and its structured data may indicate that the corresponding shooting object is located at the shooting location corresponding to the face image within the above-mentioned target shooting time period.

In one possible implementation, the selected image is the highest quality image.

In the solution provided by the embodiment of the present application, the quality of the face image is related to factors such as the illumination of the face image, the angle of the face, and the like, and can usually be expressed by a quality score.

In a second aspect, an image storage device is provided, the device comprising:

an acquisition module for the first image;

A deduplication module, configured to perform a deduplication operation on the first image and the stored image set, and is specifically used for:

Select at least one image from the first image and the stored image set to save, wherein the stored image set has one or more images, and the image in the image set is the same as the first image. An image has: the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold;

Delete unselected images.

In a possible implementation, the apparatus further includes:

The establishment module is used to obtain the second image, and if there is no third image with the same shooting object, the same shooting location, and the shooting time interval smaller than the preset threshold value as the second image, the establishment includes: an image collection of the second image.

In a possible implementation manner, the trigger condition of the deduplication operation includes:

Within a preset period of time after the first image is acquired, a fourth person who has the same subject, the same shooting location, and the shooting time interval less than the preset threshold as the first face image has not been acquired face image.

A fifth face image with the same shooting object as the first face image and a different shooting location is obtained.

It is determined that the number of face images included in the image set is greater than the target number, or,

It is determined that the size of the storage space occupied by the face images included in the image set is greater than the target storage value.

In a possible implementation manner, the deduplication module is further used for:

Modify the shooting time of the selected face image to a target shooting time period, wherein the starting time in the target shooting time period is the difference between the images in the image re-set and the shooting time of the first image. The earliest time, the end time in the target shooting time period is the latest time in the shooting time of the image in the image set and the first image;

Remove structured data and facial feature data for unselected images.

In a third aspect, a method for image storage is provided, the method comprising:

The images in the image set are acquired from the storage device, and the images in the image set and the first image have: the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold; The deduplication operation is performed on the image collection, which specifically includes: selecting an image from the image collection to save, and deleting unselected images.

In the solution provided by the embodiment of the present application, when the number of images in the image set or the occupied storage space is used as the triggering condition for the deduplication operation, the following situation may occur: the object stays at the same shooting location for a long time, and is shot by the The image capturing equipment at the location continuously captures face images, and these face images only need to save the face images with the best quality. However, due to the above trigger conditions, these face images are divided into multiple batches for deduplication. In this way, more face images will be saved. Therefore, the saved face images can also be deduplicated twice. Some images saved in the above cases meet the conditions of having the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold, these images can be considered to belong to the same image set, and the quality can be selected for these images. The best image is saved, and the remaining images can be deleted.

In a fourth aspect, an image storage device is provided, the device comprising:

an acquisition module, configured to acquire an image in an image set from a storage device, where the image in the image set and the first image have: the same shooting object, the same shooting location, and a shooting time interval less than a preset threshold;

The deduplication module is used to perform deduplication operation on the image set, and is specifically used for:

Select an image from the image collection to save;

Delete unselected images.

In a fifth aspect, a computer device is provided, the computer device includes a processor and a memory, the memory stores at least one instruction, the instruction is loaded and executed by the processor to implement the above-mentioned first aspect and The operations performed by the method for image storage described in the third aspect.

In a sixth aspect, a computer-readable storage medium is provided, the storage medium stores at least one instruction, and the instruction is loaded and executed by a processor to implement the image storage as described in the first and third aspects above the operation performed by the method.

A seventh aspect provides a computer program product that, when the computer program product is run on a computer device, causes the computer device to execute the image storage method as described in the first and third aspects above. operate.

The beneficial effects brought by the technical solutions provided in the embodiments of the present application are:

In this embodiment of the present application, after an image is acquired, de-duplication processing may be performed on the acquired image and the image in the stored image set, where the image in the image set and the currently acquired image satisfy the following conditions: have the same The shooting time, the same shooting location, and the shooting time interval is less than the preset threshold, which means that the currently acquired image and the images in the image set are multiple images continuously shot by the same object in the same shooting location in a short period of time. The images are very similar to each other. In this way, it is not necessary to save all of these images, only some images are selected to be saved, and unselected images are deleted, which can effectively save storage resources.

Description of drawings

FIG. 1 is a schematic diagram of an implementation scenario provided by an embodiment of the present application;

2 is a flowchart of a method for image storage provided by an embodiment of the present application;

3 is a schematic diagram of establishing an image set provided by an embodiment of the present application;

4 is a schematic diagram of adding a face image to an image set provided by an embodiment of the present application;

5 is a schematic diagram of a face image deduplication provided by an embodiment of the present application;

6 is a schematic structural diagram of an apparatus for image storage provided by an embodiment of the present application;

FIG. 7 is a schematic structural diagram of a computer device provided by an embodiment of the present application.

detailed description

An embodiment of the present application provides an image storage method, which can be implemented by a computer device, and the computer device has a storage function, such as a computer, a server, a network attached storage (NAS) storage system, a storage area network (storage area network, SAN) storage system, distributed storage system, cloud storage system, network video recorder (NVR), etc. Specifically, it can be an intelligent video surveillance platform (intelligent video surveillance, IVS) of Huawei. IVS can provide users with services such as face retrieval. The following is an example.

Referring to FIG. 1 , the implementation scene of the embodiment of the present application may include a front-end image capturing device and an IVS, wherein the image capturing device may be a conventional camera, or may be a smart camera that can be installed with applications, such as a software-defined camera (software defined camera). camera, SDC). Image shooting equipment can be deployed in streets, intersections and other places according to actual needs. In this embodiment of the present application, the image capturing device can send the captured images to the IVS, and the IVS deduplicates the images of the same object continuously captured by the same image capturing device in a short period of time. In this way, the storage space can be effectively saved, and the limited In the storage space, more valuable data can be stored. In addition, in the case that the above-mentioned object is a person and the image is a face image, IVS can provide the user with services such as face retrieval. , due to the deduplication operation, the total amount of face images is relatively small, which can effectively improve the efficiency of face retrieval and give users a better experience. The basis for removing duplicate images in this embodiment is that images with the same shooting object, the same shooting location, and a shooting time interval smaller than a preset threshold are considered as similar images. Duplicate image deletion is performed on these similar images. That is, only one (or multiple) of them is kept, and the rest of the images are deleted, thereby saving storage space.

The operation of removing duplicate images can have various modes of operation, such as:

Mode 1: Multiple images have been stored in the storage device (the whole of the multiple images is called an image set), and deduplication is performed between the multiple images.

Mode 2: Multiple images have been stored in the storage device (the whole of the multiple images is called an image set), and deduplication is performed between the newly received image and the multiple stored images.

Referring to FIG. 2 , a method flow of an image storage provided by an embodiment of the present application may include the following steps:

Step 201: Acquire a first image.

The acquisition location of the first image and the image set may be different, for example, the first image is acquired by the computer device from outside the storage device, and the image set is acquired by the computer device from the storage device. The first image may be acquired from the same location as the image set, for example, both acquired from a storage device by a computer device. Computer devices and storage devices can be integrated together, such as storage devices with computing capabilities.

The image may be a face image, and in the following embodiments, the image is a face image as an example for description.

In implementation, according to the different functions of the image capturing device, the method of acquiring the face image by the IVS may also be different. The following describes the method for obtaining a face image by IVS:

method one,

In the case of video capture by image capture equipment:

After the image capturing device captures the video, face recognition is performed on the video frames in the video according to the preset frame interval to obtain the face image and the scene image in the corresponding video frame. The preset frame interval may be, for example, 3 frames, 5 frames, or the like. Then, the face image in the video frame is sent to the IVS, and the IVS can directly obtain the face image.

In addition, when sending the face image of each video frame to the IVS, the image capturing device can also simultaneously send the scene image in the video frame and the corresponding shooting time and shooting location, where the shooting location is the location of the image capturing device. Alternatively, the image capturing device may not send the capturing location of the face image, but the device identification of the image capturing device.

In the case of a picture taken by an image capture device:

After the image capturing device captures a picture, face recognition is performed on the captured picture to obtain a face image and a scene image in each picture. Then, the face image in each picture is sent to the IVS, and the IVS can directly obtain the face image.

In addition, when sending the face image corresponding to each picture to the IVS, the image capturing device may also send the scene image in the picture and the corresponding shooting time and shooting location at the same time. Alternatively, the image capturing device may not send the capturing location of the face image, but the device identification of the image capturing device.

According to actual needs, IVS can process in real time, that is, after receiving the face image sent by the image capturing device, it directly acquires the face image and performs subsequent processing. Alternatively, the IVS may store the received face data first, and then acquire the face images in sequence according to the shooting time when the face image needs to be deduplicated. For example, a technician can set a deduplication period, and every time a preset deduplication period is reached, a stored face image is acquired and a deduplication operation is performed, wherein the deduplication period can be one day, three days, or the like.

Method Two,

In the case of video capture by image capture equipment:

After the image capture device captures the video, it can be directly sent to the IVS without face recognition. After receiving the video sent by the image capturing device, the IVS can perform face recognition on the video frames in the video according to the preset frame interval, and obtain the face images and scene images in the corresponding video frames.

In addition, when the image capturing device sends the video to the IVS, it can also send the shooting time information at the same time. The shooting time information may be the shooting time corresponding to each video frame in the video, or the shooting start time of the video. In the case where the shooting time information is the shooting start time of the video, after receiving the shooting start time, IVS can calculate the frame rate of the video frame that needs face recognition in the video according to the shooting start time and the frame rate of the video. filming time.

Here, the frame rate of the video may be sent by the image capturing device when sending the video, or may be stored locally by the IVS. When the IVS stores the frame rate locally, the IVS can store the frame rate and the device ID of the image capture device correspondingly, and the image capture device can carry the device ID of the machine when sending video. Correspondingly, the IVS can query the frame rate corresponding to the received device identifier according to the received device identifier and the corresponding relationship between the device identifier and the frame rate.

In addition, when sending the video and the shooting time information to the IVS, the image capturing device may also send at least one information of the shooting location of the video or the device identification of the image capturing device at the same time. In the case of sending the device identifier, the IVS may store the correspondence between the device identifier and the shooting location, and the IVS can determine the corresponding shooting location according to the device identifier.

In the case of a picture taken by an image capture device:

The image capturing device can directly send the captured pictures to the IVS without performing face recognition, and at the same time, it can also send the shooting time and shooting location of each picture.

In addition, when sending the picture, you can also send the shooting location and shooting time of the picture. After receiving the picture sent by the image capturing device, the IVS can perform face recognition on the picture to obtain the face image and scene image in the picture.

Similarly, according to actual needs, IVS can process it in real time, that is, after receiving the above data sent by the image capture device, it can directly perform face recognition to obtain a face image, or IVS can first store the received data, and then When there is a need, video frames or pictures are obtained in sequence according to the shooting time, and face recognition is performed to obtain a face image.

It should be noted here that, regardless of the first method or the second method, the acquired face images are acquired in the order of their shooting time. For example, the shooting time corresponding to the face image acquired later is later than the shooting time corresponding to the face image acquired earlier. Also, these images can be captured by the same camera or captured by different cameras.

Step 202, performing a deduplication operation on the first image and the stored image set, specifically, may include the following processing:

Select at least one image from the first image and the set of stored images to save, and delete unselected images. The stored image set contains one or more images, and the images in the image set and the first image have: the same shooting object, the same shooting location, and a shooting time interval less than a preset threshold.

The storage location of the stored image is a storage device, such as a computer, a storage server, or a NAS, a storage cloud, a distributed storage device, and the like. In particular, the stored image may be located in memory or in a non-volatile medium (eg, hard disk, solid state disk, tape).

In addition, it should be pointed out that "image collection" is a representative name for one or more images. It does not mean that a structure such as a set actually exists, nor does it mean that the executor of this step needs to perform operations such as creating a set. Of course, for the convenience of operation, images that meet the conditions can also be managed by creating sets, and images that meet the same deduplication conditions can be divided into the same set.

Considering the purpose of saving computing resources, etc., the deduplication operation may not need to be performed frequently. Therefore, the totality of images that have been received in the past and meet the deduplication (duplication) conditions can be called an image set. When a new image is received again and the conditions for performing de-duplication are met, the newly-received image and the image that has been stored in the past can be de-duplicated. Alternatively, the population (ie, the set of images) of images eligible for deduplication in the system is periodically deduplicated.

In implementation, every time a face image is obtained, it can be determined whether there is a face image stored with the same subject, the same shooting location, and the shooting time interval is less than the preset threshold as the currently obtained face image. If face images satisfying the above conditions are stored, an image set including the currently acquired face images is established. The preset threshold can be set according to actual needs, for example, can be set between 5 minutes and 10 minutes, such as 10 minutes.

As shown in FIG. 3 , a plurality of image sets have been stored, such as image set 1, image set 2 and image set 3. The currently obtained face image E, the corresponding shooting object is shooting object 4, and the shooting location is Location b, the shooting time is 10:35 on September 1, 2019, and the face image E has the same shooting object, the same shooting location, and the shooting time interval is less than a preset threshold (for example, 10 minutes) ), an image set 4 needs to be established, and the face image E is added to the image set 4.

When judging whether there is the same photographic object as the currently acquired face image, the face feature data of the currently acquired face image and the stored face features of all or part of the face images in each image set can be used The similarity calculation is performed on the data. If the similarity between the face feature data of the target face image and the face feature data of the currently obtained face image is greater than the preset similarity threshold, it can be considered that the image set where the target face image is located The face image in and the currently acquired face image correspond to the same object. The similarity calculation may be Euclidean distance calculation, cosine distance calculation, or the like.

The above part of the face image may be the face image with the highest quality in the image set, and the quality of the face image is related to the illumination of the face image, the angle of the face, and the like. There are the following methods for judging the quality of the face image: after obtaining the face image, input the pre-trained machine learning model to analyze the quality score of the face image, and the higher the quality score, the better the face image. The higher the quality. In addition, the quality of the face image can also be represented by a quality mark, different quality marks correspond to different quality levels, and the higher the quality level corresponding to the quality mark of the face image, the higher the quality of the face image.

In addition, it should be noted that whenever a face image is acquired, a pre-trained feature extraction model can be used to extract features from the currently acquired face image to obtain its corresponding face feature data, and the face The image and the facial feature data are stored correspondingly for use in calculating the similarity of the facial feature data. In addition, the pre-trained machine learning model can also be used to identify the gender, age, quality score and other data corresponding to the currently acquired face image. These data can be related to the shooting time and shooting location corresponding to the face image. The data together serve as the structured data of the face image. Correspondingly, face images, face feature data and structured data may be stored correspondingly.

After the face image is acquired, if it is determined that there is a face image stored with the same subject, the same shooting location and the shooting time interval as the currently acquired face image, and the shooting time interval is less than the preset threshold, the currently acquired face image can be stored. The face images are added to the image set where the face images satisfying the above conditions are located.

As shown in FIG. 4 , a plurality of image sets have been stored, such as image set 1, image set 2 and image set 3. The currently obtained face image E, the shooting object of which is shooting object 3, and the shooting location is site b , the shooting time is 10:40 on September 1, 2019, and the face image E has the same subject, the same shooting location, and the shooting time interval is less than the preset threshold (for example, 10 minutes). Image D, the face image D belongs to the image set 3, the face image E can be added to the image set 3.

In a possible implementation manner, a trigger condition for the deduplication operation can be set, and when the trigger condition for the deduplication operation is satisfied, the deduplication operation is performed on the face images in the image set, and at least one face image is selected to save, and no The selected face image is deleted. For the convenience of description, the face image newly added to the image set is hereinafter referred to as the first face image. There are many trigger conditions for the deduplication operation, some of which are listed below:

Trigger condition of deduplication operation 1.

Within a preset period of time after the first image is acquired, no face image is acquired that has the same shooting object, the same shooting location, and a shooting time interval smaller than the preset threshold as the first face image.

For an image set, if within a preset period of time after the first face image is acquired, the same subject, the same shooting location, and the shooting that is smaller than the preset threshold as the first face image are not acquired For face images with time interval, deduplication operation is performed on the image collection. The preset duration may be equal to the preset threshold. For example, the preset threshold may be set to 10 minutes, and the corresponding preset duration may also be set to 10 minutes, or the preset duration may be set to be slightly larger than the preset threshold.

Trigger condition for deduplication operation 2.

A face image with the same shooting object as the first face image and a different shooting location is obtained.

In this case, it means that the shooting object corresponding to the first face image has gone to a place other than the shooting location corresponding to the image set where the first face image is located, then the image set where the first face image is located can be de-duplicated operate.

Trigger condition for deduplication operation 3.

It is determined that the storage space occupied by the face images in the image set is larger than the target storage value.

The size of the storage space occupied by the face images in each image set can be detected according to the preset detection period, and if it is greater than the target storage value, the image set can be deduplicated. Among them, the target storage value can be set according to actual needs, for example, can be set to 300KB to 600KB.

Trigger condition for deduplication operation 4.

Determines the target number of face images in the image collection.

The number of face images in each image set may be detected according to a preset detection period, and if the number is greater than the preset number, a deduplication operation may be performed on the image set. The number of targets can be set according to actual needs, for example, it can be set to 10 (or 20) images, that is, deduplication is performed whenever 10 (or 20) similar images are collected. Of course, the number of targets can also be set to 2, that is, whenever an image with the same shooting object, the same shooting location and the shooting time interval less than the preset threshold is obtained as the first face image in the image set, the The image collection is deduplicated.

The third trigger condition of the deduplication operation and the fourth trigger condition of the deduplication operation can both reduce the storage pressure of the device.

Trigger condition for deduplication operation 5.

Deduplication regularly, for example, every day at 3:00 a.m., the images in the storage device are searched, and the set of images that meet the deduplication conditions (that is, the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold) multiple images) for deduplication.

It should be noted that, the trigger conditions 1, 2, and 3, or 1, 2, 4, and 5 of the above-mentioned deduplication operation can be used in combination.

In a possible implementation manner, the at least one face image selected in the above-mentioned deduplication operation is the face image with the highest quality in the image set.

In implementation, a preset number of face images may be selected for saving. Specifically, the preset number may be set according to actual requirements, and may be an integer greater than or equal to 1.

In yet another possible implementation manner, during the deduplication operation, the shooting time corresponding to the selected face image may also be modified to the target shooting time period. The start time in the target shooting time period is the earliest time in the shooting time corresponding to the face image in the image set, and the end time in the target shooting time period is the latest in the shooting time corresponding to the face image in the image set. time.

For example, as shown in Figure 5, an image set includes face image X, face image Y and face image Z, wherein:

Face image X: subject 1, shooting location a, shooting time at 10:00 on September 1, 2020, quality score 90;

Face image Y: subject 1, shooting location a, shooting time at 10:02 on September 1, 2020, quality score 87;

Face image Z: subject 1, shooting location a, shooting time at 10:06 on September 1, 2020, quality score 96.

It can be determined that the face image Z has the highest quality score, and the face image Z can be saved, and the face image X and the face image Y can be deleted. In addition, the shooting time corresponding to the face image Z can be modified to a shooting time period, wherein the starting time in the shooting time period is the earliest time among the shooting times corresponding to the face image X, the face image Y and the face image Z. : At 10:00 on September 1, 2020, the end time in the shooting time period is the latest time among the shooting times corresponding to face image X, face image Y and face image Z: September 1, 2020 10: 06.

In addition, the structured data and face feature data of unselected face images can also be deleted.

In another possible implementation manner, when the trigger condition for the deduplication operation includes the trigger condition 3 of the deduplication operation or the trigger condition 4 of the deduplication operation, it may happen that the object stays at the same shooting location for a long time, The face images are continuously captured by the image capturing equipment at the shooting location, and these face images only need to save the face images with the best quality. Fourth, these face images are divided into multiple batches for deduplication, which will save more face images. Therefore, the saved face images can also be deduplicated twice. Correspondingly, processing It can be as follows: in the two-person face images with the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold, the higher-quality face images are retained, and the lower-quality face images are deleted. , and modify the start time of the shooting time period corresponding to the high-quality face image to the earliest shooting time in the shooting time period corresponding to the above-mentioned two-face image, and modify the end time to the shooting time period corresponding to the above-mentioned two-face image The latest shooting time in .

For example, the following face images are stored:

Face image 1: subject 1, shooting location a, shooting time period 10:00 on September 1, 2020 to 10:15 on September 1, 2020, quality score 90;

Face image 2: subject 1, shooting location a, shooting time period from 10:17 on September 1, 2020 to 10:30 on September 1, 2020, quality score 85;

Face image 3: subject 1, shooting location a, shooting time period from 10:35 on September 1, 2020 to 10:50 on September 1, 2020, with a quality score of 95.

First, it is determined that the photographing objects and locations corresponding to face image 1 and face image 2 are the same, and the minimum time interval of the corresponding photographing time period is 2 minutes less than a preset threshold (for example, 10 minutes), and then it is determined that face image 1 The quality score is greater than the quality score of face image 2, then face image 1 is saved, and face image 2 is deleted. In addition, the start time of the shooting time period corresponding to face image 1 is modified to the earliest shooting time in the shooting time period corresponding to face image 1 and face image 2: 10:00 on September 1, 2020, the face The end time of the shooting time period corresponding to image 1 is modified to the latest shooting time in the shooting time period corresponding to face image 1 and face image 2: 10:30 on September 1, 2020. After modification, the relevant information of face image 1 is as follows:

Face image 1: subject 1, shooting location a, shooting time period from 10:00 on September 1, 2020 to 10:30 on September 1, 2020, with a quality score of 90.

Then, it is determined that the subject and the shooting location corresponding to the modified face 1 and the face image 3 are the same, and the minimum time interval of the corresponding shooting time period is 5 minutes less than the above-mentioned preset threshold, and it is determined that the face image 3 If the quality score is greater than the quality score of face image 1, face image 3 is retained and face image 1 is deleted. In addition, the start time of the shooting time period corresponding to face image 3 is modified to the earliest shooting time in the shooting time period corresponding to face image 1 and face image 3: 10:00 on September 1, 2020, the face The end time of the shooting time period corresponding to image 3 is modified to the latest shooting time in the shooting time period corresponding to face image 1 and face image 3: 10:50 on September 1, 2020. After modification, the relevant information of face image 3 is as follows:

Face image 3: subject 1, shooting location a, shooting time period from 10:00 on September 1, 2020 to 10:50 on September 1, 2020, with a quality score of 95.

It can be seen that through the above processing, the face image can be further de-duplicated, so as to achieve the purpose of retaining only the highest-quality face image of the same object shot at the same shooting location in a short period of time. storage.

Based on the same technical concept, an embodiment of the present application also provides a device for image storage, as shown in Figure 6, the device includes:

an acquisition module 610, configured to acquire the first image, and specifically can implement the acquisition function mentioned in the above step 201, as well as other implicit steps;

The deduplication module 620 is used to perform deduplication operation on the first image and the stored image set, and is specifically used for:

Delete unselected images. Specifically, the deduplication module 620 can implement the deduplication function mentioned in the above step 202, as well as other implicit steps.

In a possible implementation manner, the apparatus further includes:

The establishment module 630 is used to obtain the second image, if there is no third image with the same shooting object, the same shooting location, and the shooting time interval smaller than the preset threshold as the second image, then establish An image collection including the second image.

In a possible implementation manner, the deduplication module 620 is further configured to:

Remove structured data and facial feature data for unselected images.

In a possible implementation manner, the selected image is the image with the highest quality, such as the one image with the highest quality, or the two images with the highest quality.

In addition, the acquiring module 610 is further configured to acquire images in the image set from the storage device, where the images in the image set and the first image have: the same shooting object, the same shooting location, and shooting less than a preset threshold time interval;

The deduplication module 620 is further configured to perform deduplication operations on the first image and the stored image set, specifically including:

Select an image from the image collection to save;

Delete unselected images.

It should be noted that: when the image storage device provided in the above embodiments is used for image storage, only the division of the above functional modules is used as an example for illustration. That is, the internal structure of the computer equipment is divided into different functional modules to complete all or part of the functions described above. In addition, the image storage apparatus and the image storage method embodiments provided by the above embodiments belong to the same concept, and the specific implementation process thereof is detailed in the method embodiments, which will not be repeated here.

FIG. 7 is a schematic structural diagram of a computer device provided by an embodiment of the present application. The computer device may be the IVS in the above method embodiment. The computer device 700 may vary greatly due to different configurations or performances, and may include a or more than one processor (central processing units, CPU) 701 and one or more than one memory 702, wherein, the memory 702 stores at least one instruction, the at least one instruction is loaded and executed by the processor 701 to The methods provided by the foregoing method embodiments are implemented. Of course, the server may also have components such as a wired or wireless network interface, a keyboard, and an input/output interface for input and output, and the server may also include other components for implementing device functions, which will not be described here.

In the above embodiments, it may be implemented in whole or in part by software, hardware, firmware or any combination thereof, and when implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions, and when the computer program instructions are loaded and executed on a device, all or part of the processes or functions described in the embodiments of the present application are generated. The computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be downloaded from a website site, computer, server, or data center Transmission to another website site, computer, server, or data center by wire (eg, coaxial cable, optical fiber, digital subscriber line) or wireless (eg, infrared, wireless, microwave, etc.). The computer-readable storage medium may be any available medium that the device can access, or a data storage device such as a server, data center, or the like that includes an integration of one or more available media. The usable medium may be a magnetic medium (such as a floppy disk, a hard disk, and a magnetic tape, etc.), an optical medium (such as a digital video disk (Digital Video Disk, DVD), etc.), or a semiconductor medium (such as a solid-state disk, etc.).

All or part of the steps in the above-mentioned embodiments can be completed by hardware, or can be completed by instructing relevant hardware through a program, and the program can be stored in a computer-readable storage medium, and the above-mentioned storage medium can be only a computer-readable storage medium. Read memory, disk or CD, etc.

The above description is only an embodiment of the present application, and is not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made should be included within the protection scope of the present application.

Claims

A method for image storage, characterized in that the method comprises:

get the first image;

Perform a deduplication operation on the first image and the stored image set, specifically including:

Select at least one image from the first image and the stored image set to save, wherein the stored image set has one or more images, and the image in the image set is the same as the first image. An image has: the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold;

Delete unselected images.
The method according to claim 1, wherein the method further comprises:

get the second image;

If there is no third image having the same photographing subject, the same photographing location, and a photographing time interval smaller than the preset threshold as the second image, an image set including the second image is established.
The method according to claim 1, wherein the triggering condition of the deduplication operation comprises:

Within a preset period of time after the first image is acquired, a fourth person who has the same subject, the same shooting location, and the shooting time interval less than the preset threshold as the first face image has not been acquired face image.
The method according to claim 1, wherein the triggering condition of the deduplication operation comprises:

A fifth face image with the same shooting object as the first face image and a different shooting location is obtained.
The method according to claim 1, wherein the triggering condition of the deduplication operation comprises:

It is determined that the number of face images included in the image set is greater than the target number, or,

It is determined that the size of the storage space occupied by the face images included in the image set is greater than the target storage value.
The method according to any one of claims 1-5, wherein the method further comprises:

Modify the shooting time of the selected face image to a target shooting time period, wherein the starting time in the target shooting time period is the difference between the images in the image re-set and the shooting time of the first image. The earliest time, the end time in the target shooting time period is the latest time in the shooting time of the image in the image set and the first image;

Remove structured data and facial feature data for unselected images.
The method according to any one of claims 1-6, wherein the selected image is the image of the highest quality.
An image storage device, characterized in that the device comprises:

an acquisition module for the first image;

A deduplication module, configured to perform a deduplication operation on the first image and the stored image set, and is specifically used for:

Select at least one image from the first image and the stored image set to save, wherein the stored image set has one or more images, and the image in the image set is the same as the first image. An image has: the same shooting object, the same shooting location, and the shooting time interval less than the preset threshold;

Delete unselected images.
The apparatus according to claim 8, wherein the apparatus further comprises:

The establishment module is used to obtain the second image, and if there is no third image with the same shooting object, the same shooting location, and the shooting time interval smaller than the preset threshold value as the second image, the establishment includes: an image collection of the second image.
The device according to claim 8, wherein the trigger condition of the deduplication operation comprises:

Within a preset period of time after the first image is acquired, a fourth person who has the same subject, the same shooting location, and the shooting time interval less than the preset threshold as the first face image has not been acquired face image.
The device according to claim 8, wherein the trigger condition of the deduplication operation comprises:

A fifth face image with the same shooting object as the first face image and a different shooting location is obtained.
The device according to claim 8, wherein the trigger condition of the deduplication operation comprises:

It is determined that the number of face images included in the image set is greater than the target number, or,

It is determined that the size of the storage space occupied by the face images included in the image set is greater than the target storage value.
The device according to any one of claims 8-12, wherein the deduplication module is further used for:

Modify the shooting time of the selected face image to a target shooting time period, wherein the starting time in the target shooting time period is the difference between the images in the image re-set and the shooting time of the first image. The earliest time, the end time in the target shooting time period is the latest time in the shooting time of the image in the image set and the first image;

Remove structured data and facial feature data for unselected images.
14. The apparatus of any one of claims 8-13, wherein the selected image is the highest quality image.
A computer device, characterized in that the computer device comprises a processor and a memory, the memory stores at least one instruction, and the instruction is loaded and executed by the processor to realize the invention as claimed in claim 1 to claim 7 The operations performed by the image storage method described in any one of the above.
A computer-readable storage medium, wherein at least one instruction is stored in the storage medium, and the instruction is loaded and executed by a processor to realize the image according to any one of claims 1 to 7 The operation performed by the stored method.
A method for image storage, characterized in that the method comprises:

Acquiring images in the image set from the storage device, where the images in the image set and the first image have: the same shooting object, the same shooting location, and a shooting time interval less than a preset threshold;

Perform a deduplication operation on the first image and the stored image set, specifically including:

Select an image from the image collection to save;

Delete unselected images.
An image storage device, characterized in that the device comprises:

an acquisition module, configured to acquire images in the image set from the storage device, where the images in the image set and the first image have: the same shooting object, the same shooting location, and a shooting time interval less than a preset threshold;

The deduplication module is used to perform deduplication operation on the image set, and is specifically used for:

Select an image from the image collection to save;

Delete unselected images.