WO2017016452A1 - 一种图片处理装置及其方法 - Google Patents

一种图片处理装置及其方法 Download PDF

Info

Publication number
WO2017016452A1
WO2017016452A1 PCT/CN2016/091060 CN2016091060W WO2017016452A1 WO 2017016452 A1 WO2017016452 A1 WO 2017016452A1 CN 2016091060 W CN2016091060 W CN 2016091060W WO 2017016452 A1 WO2017016452 A1 WO 2017016452A1
Authority
WO
WIPO (PCT)
Prior art keywords
picture
quality value
module
social
value
Prior art date
Application number
PCT/CN2016/091060
Other languages
English (en)
French (fr)
Inventor
杜守富
王颖
Original Assignee
中兴通讯股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2017016452A1 publication Critical patent/WO2017016452A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q50/00—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking

Definitions

  • the present application relates to, but is not limited to, a media processing technology of a smart device, and more particularly to a picture processing device and a method thereof.
  • the related technologies are based on image recognition technology, which performs invalid image filtering, identifies a certain type of image, and processes it, mainly for garbage pictures (such as spam or yellow pictures containing advertisement pictures).
  • the picture has distinct features.
  • the disadvantage of the related art is that it directly depends on the accuracy of the image recognition algorithm, and the image recognition technology is generally affected by the image noise, and the accuracy is difficult to improve.
  • these image processing techniques are computationally intensive and require high performance on the terminal.
  • the invention patent application with the application number CN201410392590.7 proposes a method for not storing the photographed picture with poor automatic filtering effect, mainly relies on the image processing technology, and whether the picture is closed, whether the virtual focus or the like is not effective. The indicator is discriminated.
  • the related art image filtering method based on image recognition technology has its inherent defects, and needs to rely on the accuracy of the image recognition itself, and is difficult to identify the semantic layer.
  • This paper provides a picture processing device and a processing method thereof that can effectively improve the accuracy of garbage picture recognition.
  • a picture processing device includes: a difference picture detection module, a local program monitoring module, and a statistics center module; the difference picture detection module is configured to: identify a feature of the stored picture, according to the picture The feature acquires a content quality value of the picture and sends the content quality value of the picture to a statistical center module; the local program monitoring module is configured to: record a usage parameter value of the picture in the application, according to the picture Obtaining a first social quality value of the picture using a parameter value and transmitting the first social quality value of the picture to a statistical center module; the statistics center module is configured to: receive a content quality value of the picture and a first social quality And calculating a picture quality of the picture according to the content quality value of the picture and the first social quality value.
  • the foregoing statistical center module includes a weight management sub-module and a comprehensive calculation sub-module; the weight management sub-module is configured to: respectively assign weights to the content quality value and the first social quality value of the image; and the comprehensive calculation sub-module, The method is configured to: calculate a picture quality value of the picture according to the content quality value of the picture and the corresponding weight and the first social quality value and the corresponding weight.
  • the foregoing statistics center module further includes a network communication sub-module and a picture repetition detection sub-module; the network communication sub-module is configured to: identify other devices synchronized with the local device, and receive the device identifier sent by the other device and the corresponding picture quality. Value; picture repeat detection sub-module, set to: detect whether there is a picture stored on the local device repeatedly with other devices; the comprehensive calculation sub-module is also set to: when the local device and other devices have duplicate pictures, update the picture of the duplicate picture Quality value.
  • the foregoing network communication sub-module is further configured to: when the local device and the other device have duplicate pictures, notify other devices having the repeated pictures to update the picture quality value of the duplicate picture.
  • the foregoing statistics center module further includes a deletion submodule, configured to: determine a picture to be deleted according to the picture quality value of the picture.
  • the foregoing deleting sub-module includes: a marking unit, configured to: mark a picture lower than the picture quality threshold as a picture to be deleted according to a preset picture quality threshold; and delete the unit, set to: when the deletion is received When the command is executed, the picture to be deleted is deleted.
  • the foregoing apparatus further includes: a social relationship monitoring module, configured to: obtain a second social quality value according to the communication parameter value of the contact, and send the second social quality value to the statistical center module; and the statistical center module is further configured to: receive the social relationship monitoring a second social quality value sent by the module, wherein the weight management sub-module is further configured to: assign a weight to the second social quality value of the stored picture; the comprehensive calculation sub-module is further configured to: according to the content quality value and the corresponding weight, The picture quality of the stored picture is calculated by a social quality value and a corresponding weight, and a second social quality value and a corresponding weight.
  • a social relationship monitoring module configured to: obtain a second social quality value according to the communication parameter value of the contact, and send the second social quality value to the statistical center module
  • the statistical center module is further configured to: receive the social relationship monitoring a second social quality value sent by the module
  • the weight management sub-module is further configured to: assign a weight to the second social quality value of
  • the foregoing social relationship monitoring module includes a communication sub-module and a face recognition module; the communication sub-module is configured to: record a communication parameter value of the contact; and the face detection module is configured to: detect whether the communication exists in the image a contact picture in the record; when it is detected that there is a picture of the contact in the communication record in the stored picture, the second social quality value is obtained according to the communication parameter value of the record contact, and the second social quality of the picture is obtained The value is sent to the statistics center module;
  • a method for processing a picture comprising: identifying a ⁇ feature of a stored picture, obtaining a content quality value of the picture according to a ⁇ feature of the picture; recording a usage parameter value of the picture in the application, according to the The first social quality value is obtained by using the parameter value; and the picture quality of the stored picture is calculated according to the content quality value of the picture and the first social quality value.
  • the foregoing calculating, according to the content quality value of the picture and the first social quality value, the picture quality of the stored picture comprising: respectively assigning a weight to the content quality value and the first social quality value of the picture; and according to the content quality of the stored picture
  • the picture quality value of the picture is calculated by the value and the corresponding weight and the first social quality value and the corresponding weight.
  • the foregoing method further includes: identifying another device that is synchronized, receiving the device identifier sent by the other device and the corresponding picture quality value; detecting whether the local device has a duplicate image stored with the other device; and when the local device has the other device Repeat the picture to update the picture quality value of the duplicate picture.
  • the foregoing method further includes: when the local device has the repeatedly stored picture, notifying other devices having the repeated picture to update the quality value of the duplicate picture.
  • the foregoing method further includes: determining a picture to be deleted according to the picture quality value of the picture.
  • determining, according to the picture quality value of the picture, the picture to be deleted includes: marking, according to a preset picture quality threshold, a picture that is lower than the picture quality threshold as a picture to be deleted; when receiving the delete command When the picture to be deleted is deleted.
  • the foregoing method further includes: acquiring a second social quality of the picture according to a communication parameter value of the contact; assigning a weight to the second social quality value of the stored picture; and according to the content quality value and the corresponding weight, the first The social quality value and the corresponding weight, and the second social quality value and the corresponding weights calculate the picture quality of the stored picture.
  • the foregoing obtaining the second social quality according to the communication parameter value of the contact includes: recording a communication parameter value of the contact; detecting whether a contact picture in the communication record exists in the picture; and detecting the storage When there is a picture of the contact in the communication record in the picture, the second social quality value of the picture is obtained according to the communication parameter value of the record contact.
  • the foregoing communication parameter value includes at least one of the following: the number of communications this week, the number of communications this month, the sum of the number of communications in the current year, and the total length of communication.
  • the usage parameter value of the foregoing picture in the application includes at least one of the following: the number of times of access in the application and the length of time of the access.
  • the foregoing ⁇ picture feature includes at least one of: a picture virtual focus feature and a picture closed eye feature.
  • a computer readable storage medium storing computer executable instructions that, when executed by a processor, implement the above method.
  • the storage of the device based on the embodiment of the present invention is limited, it takes a certain time to reach the storage threshold. During this period, multiple user operations may occur for the picture, and the operation of the picture is highly correlated with the quality of the picture itself. Sex, if the user browses the picture, it will definitely stay longer for a good picture, and even upload the pictures that are considered better to Weibo, WeChat, QQ and other social software. Based on this fact, this article collects users.
  • the operation behavior of the picture proposes a large number of image processing methods mainly stored in the device, which can increase the judgment of different quality values on the basis of the existing picture recognition, thereby realizing the accuracy of the garbage picture recognition.
  • FIG. 1 is a schematic diagram of a picture processing apparatus according to an embodiment of the present invention.
  • FIG. 2 is a schematic structural diagram of synchronizing processing performed by a picture processing apparatus on different devices according to an embodiment of the present invention
  • FIG. 3 is a schematic flowchart of a picture processing method according to an embodiment of the present invention.
  • FIG. 1 is a schematic diagram of a picture processing apparatus according to an embodiment of the present invention.
  • the image processing module 11 includes a difference picture detecting module 11, a local program monitoring module 12, and a statistics center module 13.
  • the difference picture detecting module 12 is configured to: identify a ⁇ feature of the stored picture, perform content quality value evaluation on the ⁇ feature of the picture, obtain a content quality value of the picture, and send the content quality value of the picture to the statistical center module. 13.
  • the local program monitoring module 12 is configured to: record a usage parameter value of the picture in the application, perform a first social quality value evaluation on the picture according to the use parameter value, and obtain a first social quality value of the picture and the picture
  • the corresponding first social quality value is sent to the statistics center module 13.
  • the greater the first social quality value the higher the quality of the picture, and thus the first social quality value is a positive value.
  • the statistics center module 13 is configured to: receive a content quality value of the picture and a first social quality value, and calculate a picture quality of the picture according to the content quality value of the picture and the first social quality value.
  • the value of the usage parameter of the picture in the application includes at least one of the following: in the application The number of times the program is accessed and the length of the access, wherein the application can be a photo program, a photo browsing program, a microblog, a WeChat, a QQ, a photo editor, and the like.
  • the ⁇ picture feature includes at least one of: a picture virtual focus feature and a picture closed eye feature; other ⁇ features may also be red eye, exposure, etc., and are not limited only to the above ⁇ feature; wherein the virtual focus feature may be based on a difference operator method The best curve fitting, prewitt operator algorithm and other algorithms calculated that the closed eye feature can be derived based on the texture segmentation method and the adaboost algorithm.
  • the image processing apparatus of the embodiment of the present invention can be applied to terminal devices such as mobile phones, tablets, and personal computers, but is not limited to the above devices.
  • the statistics center module 13 may include a weight management sub-module 131 and a comprehensive calculation sub-module 132.
  • the weight management sub-module 131 is configured to: respectively assign a weight to the content quality value and the first social quality value of the picture; the comprehensive calculation sub-module 132 And being set to: calculate a picture quality value of the picture according to the content quality value of the picture and the corresponding weight and the first social quality value and the corresponding weight. It can also be set to increase the repeatedly stored picture quality value when the local device has a duplicate stored picture.
  • the statistics center module 13 may further include a network communication sub-module 133 and a picture repetition detection sub-module 134.
  • the network communication sub-module 133 is configured to: identify other devices synchronized with the local device, and receive device identifications and corresponding picture quality sent by other devices.
  • the value repeating detection sub-module 134 is configured to: detect whether there is a picture stored on the local device repeatedly with other devices.
  • the integrated calculation sub-module 132 is further configured to: when the local device has a repeatedly stored picture, update the repeatedly stored picture quality value, and the embodiment adopts an increased quality value.
  • the networked communication sub-module 133 may be further configured to notify other devices having duplicate stored pictures when the local device has a repeatedly stored picture.
  • FIG. 2 is a schematic structural diagram of synchronizing processing performed by a picture processing apparatus on different devices according to an embodiment of the present invention.
  • Elections can be used to elect terminals with strong processing power for image quality value coordination, such as computers, computers to receive image quality values already calculated by other devices, and to calculate unique feature values of images according to certain algorithms, similar to picture DNA, try to The only one is recorded in the record of the computer picture at the same time, and the value of the repeatedly stored picture quality is increased.
  • Other devices may be mobile phones, computers, tablets, etc., and when receiving repeated images stored by the computer, they also respond to recalculating the picture quality of the pictures on each device.
  • the statistics center module 13 may further include a deletion sub-module 135, including a marking unit, configured to: mark a picture lower than the picture quality threshold as a picture to be deleted according to a preset picture quality threshold; and delete the unit, set as: When the delete command is received, the picture to be deleted is deleted.
  • a deletion sub-module 135 including a marking unit, configured to: mark a picture lower than the picture quality threshold as a picture to be deleted according to a preset picture quality threshold; and delete the unit, set as: When the delete command is received, the picture to be deleted is deleted.
  • the image processing apparatus may further include a social relationship monitoring module 14 configured to: perform a second social quality assessment on the image to obtain a second social quality value according to the communication parameter value of the contact.
  • the second value of the second social quality value indicates that the quality of the picture is higher, so the second social quality value is a positive value.
  • the social relationship monitoring module 14 includes a communication sub-module 141 and a face recognition module 142.
  • the communication sub-module 141 is configured to: record a communication parameter value of the contact, and the face recognition module 142 is configured to: detect whether there is communication in the picture.
  • the statistics center module 13 is further configured to: receive the second social quality value sent by the social relationship monitoring module 14, wherein the weight management sub-module 131 further The method is configured to: assign a weight to the second social quality value of the stored picture; the comprehensive calculation sub-module 132 is further configured to: according to the content quality value and the corresponding weight, the first social quality value and the corresponding weight, and the second social quality value And the corresponding weights calculate the picture quality of the stored picture.
  • the communication parameter value may include at least one of the following: the number of communications this week, the number of communications this month, the sum of the number of communications this year, and the total length of communication.
  • Table 1 is a schematic diagram of an algorithm of a picture processing apparatus according to an embodiment of the present invention. The calculation is as follows:
  • the image processing apparatus of the embodiment of the invention can add the image to the identified image, and has the function of manual manual checking. If the image detected by the image processing apparatus is found to be inaccurate, the weight adjustment can be performed to obtain a more accurate filtering result.
  • the quality values and weights of the above images can be default values or manually set by the user.
  • FIG. 3 is a schematic flowchart of a picture processing method according to an embodiment of the present invention. As shown in FIG. 3, the method includes:
  • S101 Identify a ⁇ feature of the stored picture, and perform content quality value evaluation on the ⁇ feature of the picture to obtain a content quality value of the picture;
  • S102 Record a usage parameter value of the picture in an application, and perform a first social quality value evaluation on the picture according to the usage parameter value to obtain a first social quality value of the picture; optionally, the first social quality The larger the value, the higher the quality of the picture, so the first social quality value is positive.
  • S103 Calculate a picture quality of the stored picture according to the content quality value of the picture and the first social quality value.
  • the method includes: assigning a weight to the content quality value and the first social quality value of the picture respectively; and according to the content quality value of the stored picture and the corresponding weight and the first social quality value and the corresponding right Recalculate the picture quality value of the picture.
  • the method may further include: S104: determining a picture to be deleted according to a picture quality value of the stored picture, where the method includes marking, according to a preset picture quality threshold, a picture that is lower than the picture quality threshold as a picture to be deleted; when receiving When the command is deleted, the picture to be deleted is deleted.
  • the usage parameter value of the picture in the application includes at least one of the following: the number of times of access in the application and the length of time of the access, wherein the application may be a photo program, a picture browsing program, a microblog, a WeChat , QQ, photo editing, etc.
  • the ⁇ picture feature includes at least one of: a picture virtual focus feature and a picture closed eye feature; other ⁇ features may also be red eye, exposure, etc., and are not limited only to the above ⁇ feature; wherein the virtual focus feature may be based on a difference operator method The best curve fitting, prewitt operator algorithm and other algorithms calculated that the closed eye feature can be derived based on the texture segmentation method and the adaboost algorithm.
  • the image processing method of the embodiment of the present invention can be applied to terminal devices such as mobile phones, tablets, and personal computers, but is not limited to the above devices.
  • the image processing method further includes: performing a second social quality assessment on the image according to the communication parameter value of the contact; assigning a weight to the second social quality value, according to the content quality value of the image, the first social quality The value, and the second social quality value, calculate the picture quality of the stored picture.
  • the greater the second social quality value the higher the quality of the picture, so the value is a positive value.
  • the obtaining the second social quality value by performing the second social quality assessment on the image according to the communication parameter value of the contact may include: monitoring a contact record parameter value of the contact, and detecting whether there is a contact in the communication record in the image. a person picture; when it is detected that the picture of the contact in the communication record exists in the stored picture, performing a second social quality value evaluation on the picture according to the communication record parameter value of the contact to obtain the second social quality value.
  • the image processing method further includes: identifying other devices that are synchronized, receiving other device sending device identifiers and corresponding image quality values; detecting whether the local device has duplicate images stored with other devices; and when the local device has duplicate storage The picture quality value of the repeated picture is updated, and when the local device has the repeatedly stored picture, the other device that has the repeatedly stored picture is notified to update the picture quality value of the duplicate picture;
  • Elections can be used to elect terminals with strong processing power for image quality value coordination.
  • computers and computers receive image quality values that have been calculated by other devices, and calculate unique eigenvalues of images according to a certain algorithm, for example, pixel mean value.
  • Greek algorithm first cut and shrink the picture
  • To the standard 4 8*8 pictures compare the pixels of each picture with the pixel mean. If each pixel is greater than the average, the pixel base is set to 1, less than 0, then an 8*8 picture has 64.
  • An overall picture consists of 256 strings representing the feature values of the picture. Similar to the picture DNA, try to be unique, and record it in the record of the computer picture, and increase the value of the repeatedly stored picture quality.
  • Other devices also respond to recalculate the picture quality of the pictures on each device when receiving the repeatedly stored pictures sent by the computer.
  • the image processing method in the embodiment of the present invention can add the image to the identified image, and has the function of manual manual checking. If the image detected by the image processing method is found to be inaccurate, the weight adjustment can be performed to obtain a more accurate filtering result.
  • the embodiment of the invention further provides a computer readable storage medium storing computer executable instructions, which are implemented when the computer executable instructions are executed by the processor.
  • all or part of the steps of the above embodiments may also be implemented by using an integrated circuit. These steps may be separately fabricated into individual integrated circuit modules, or multiple modules or steps may be fabricated into a single integrated circuit module. achieve.
  • the devices/function modules/functional units in the above embodiments may be implemented by a general-purpose computing device, which may be centralized on a single computing device or distributed over a network of multiple computing devices.
  • the device/function module/functional unit in the above embodiment When the device/function module/functional unit in the above embodiment is implemented in the form of a software function module and sold or used as a stand-alone product, it can be stored in a computer readable storage medium.
  • the above mentioned computer readable storage medium may be a read only memory, a magnetic disk or an optical disk or the like.

Abstract

一种图片处理装置及其方法，所述装置包括：差质图片检测模块（11）、本地程序监控模块（12）及统计中心模块（13）；差质图片检测模块（11），设置为：获取所述图片的内容质量值并将所述图片的内容质量值发送至统计中心模块（13）；本地程序监控模块（12），设置为：获取所述图片的第一社交质量值并将所述图片的第一社交质量值发送至统计中心模块（13）；统计中心模块（13），设置为：根据所述图片的内容质量值和第一社交质量值计算所述图片的图片质量。

Description

一种图片处理装置及其方法 技术领域
本申请涉及但不限于智能设备的媒体处理技术，尤指一种图片处理装置及其方法。
背景技术
随着手持移动设备的大量普及，如iphone和ipad等的智能设备的流行，在设备上产生了大量的图片资源。如手机拍照，微博、微信等应用，无时无刻不在产生大量的图片。积累到一定的时间，终端存储会不够用，即使PC机等大容量存储设备，也很快被这些高清保留的图片进行充满。这时如果想在大量图片中检索出不希望保存的图片，从而进行永久删除，对用户就有着巨大的工作量。因此，如何有效的管理大量图片，如在移动设备存储有效的情况，如何快速甄别垃圾图片，从而能有效的删除垃圾图片，来节省存储空间成为越来越重要的一个图片管理功能。
相关技术中，大部分是基于图像识别技术，进行无效图片过滤，识别出某种特定类型的图片，进行处理，主要是针对垃圾图片(如包含广告图片的垃圾邮件或黄色图片等)，该类图片具有明显的特征。相关技术的缺点是，直接依赖图像识别算法的精确度，而且图像识别技术一般受图片噪声影响严重，精度很难提高。而且，这些图片处理技术计算量大，对终端性能要求较高。例如申请号为CN201410392590.7的发明专利申请提出一种自动过滤效果不佳的拍照图片不进行存储的方法，主要依赖图片处理技术，查看图片是否闭眼，是否虚焦等图片效果不佳的主要指标进行判别。该方法，因为图片技术的本身缺点，识别精度有待提高，另外对于图片中包含的语义层的信息无法判断，造成误判。比如，一张有人闭眼的图片，但图片是人像合照，且只有这一张图片，比较珍贵，在拍摄者看来具有保留价值的图片，也被误判。
从上面可以看出，相关技术基于图像识别技术的图片过滤方法，有其固有的缺陷，需要依赖图片识别本身精度，对语义层的较难识别。
发明内容
以下是对本文详细描述的主题的概述。本概述并非是为了限制权利要求的保护范围。
本文提供一种能够有效提高对垃圾图片识别的准确率的图片处理装置及其处理方法。
一种图片处理装置，所述装置包括：差质图片检测模块、本地程序监控模块、及统计中心模块；差质图片检测模块，设置为：识别存储的图片的瑕疵特征，根据所述图片的瑕疵特征获取所述图片的内容质量值并将所述图片的内容质量值发送至统计中心模块；本地程序监控模块，设置为：记录所述图片在应用程序中的使用参数值，根据所述图片的使用参数值获取所述图片的第一社交质量值并将所述图片的第一社交质量值发送至统计中心模块；统计中心模块，设置为：接收所述图片的内容质量值和第一社交质量值，根据所述图片的内容质量值和第一社交质量值计算所述图片的图片质量。
可选地，前述统计中心模块包括权重管理子模块和综合计算子模块；权重管理子模块，设置为：对所述图片的内容质量值和第一社交质量值分别分配权重；综合计算子模块，设置为：根据图片的内容质量值及对应的权重和第一社交质量值及对应的权重计算图片的图片质量值。
可选地，前述统计中心模块还包括联网通讯子模块和图片重复检测子模块；联网通讯子模块，设置为：识别与本地设备同步的其他设备，接收其他设备发送的设备标识及对应的图片质量值；图片重复检测子模块，设置为：检测本地设备上是否有与其他设备重复存储的图片；综合计算子模块，还设置为：当本地设备与其他设备具有重复图片时，更新重复图片的图片质量值。
可选地，前述联网通讯子模块，还设置为：当本地设备与其他设备具有重复图片时，通知具有重复图片的其他设备更新重复图片的图片质量值。
可选地，前述统计中心模块还包括删除子模块，设置为：根据所述图片的图片质量值，确定待删除的图片。
可选地，前述删除子模块包括：标记单元，设置为：根据预设的图片质量阈值，将低于所述图片质量阈值的图片标记为待删除图片；删除单元，设置为：当接收到删除命令时，删除所述待删除图片。
可选地，前述装置还包括：社交关系监控模块，设置为：根据联系人的通讯参数值，获得第二社交质量值并发送至统计中心模块；统计中心模块，还设置为：接收社交关系监控模块发送的第二社交质量值，其中，权重管理子模块还设置为：为存储图片的第二社交质量值分配权重；综合计算子模块，还设置为：根据内容质量值及对应的权重、第一社交质量值及对应的权重、及第二社交质量值及对应的权重计算存储图片的图片质量。
可选地，前述社交关系监控模块包括通讯录子模块和人脸识别模块；通讯录子模块，设置为：记录联系人的通讯参数值；人脸检测模块，设置为：检测图片中是否存在通讯记录中的联系人图片；当检测到存储的图片中存在通讯记录中的联系人的图片时，根据记录联系人的通讯参数值获取第二社交质量值，并将所述图片的第二社交质量值发送至统计中心模块；
一种图片处理方法，所述方法包括：识别存储的图片的瑕疵特征，根据所述图片的瑕疵特征获取所述图片的内容质量值；记录所述图片在应用程序中的使用参数值，根据所述使用参数值获取第一社交质量值；根据所述图片的内容质量值和第一社交质量值计算存储图片的图片质量。
可选地，前述根据图片的内容质量值和第一社交质量值计算存储图片的图片质量，包括：对所述图片的内容质量值和第一社交质量值分别分配权重；根据存储图片的内容质量值及对应的权重和第一社交质量值及对应的权重计算所述图片的图片质量值。
可选地，前述方法还包括：识别同步的其他设备，接收其他设备发送的设备标识及对应的图片质量值；检测本地设备上是否有与其他设备重复存储的图片；当本地设备具有与其他设备重复的图片，更新重复的图片的图片质量值。
可选地，前述方法还包括：当本地设备具有重复存储的图片时，通知具有重复图片的其他设备更新重复图片的质量值。
可选地，前述方法还包括：根据所述图片的图片质量值，确定待删除的图片。
可选地，前述根据所述图片的图片质量值，确定待删除的图片包括：根据预设的图片质量阈值，将低于所述图片质量阈值的图片标记为待删除图片；当接收到删除命令时，删除所述待删除图片。
可选地，前述方法还包括：根据联系人的通讯参数值，获取所述图片的第二社交质量；对存储图片的第二社交质量值分配权重；根据内容质量值及对应的权重、第一社交质量值及对应的权重、及第二社交质量值及对应的权重计算存储图片的图片质量。
可选地，前述根据联系人的通讯参数值，获取所述图片进行第二社交质量，包括：记录联系人的通讯参数值；检测图片中是否存在通讯记录中的联系人图片；当检测到存储的图片中存在通讯记录中的联系人的图片时，根据记录联系人的通讯参数值获取所述图片的第二社交质量值。
可选地，前述通讯参数值包括如下至少之一：本周通讯次数、本月通讯次数、本年通讯次数之和、及通讯总时长。
可选地，前述图片在应用程序中的使用参数值包括如下至少之一：在应用程序中被访问的次数和访问的时间长度。
可选地，前述瑕疵图片特征包括至少其中之一：图片虚焦特征和图片闭眼特征。
一种计算机可读存储介质，存储有计算机可执行指令，所述计算机可执行指令被处理器执行时实现上述方法。
本发明实施例基于设备的存储虽然有限，但达到存储的阈值是需要一定的时间，在此期间针对图片可能会发生多种用户操作，而这些针对图片的操作跟图片本身质量有极强的相关性，如用户浏览图片时，针对好的图片肯定会停留的时间较长，甚至会把认为较好的图片上传到微博，微信，QQ等社交软件，基于这样的事实，本文通过采集用户对图片的操作行为，作为权重指标提出一种主要针对设备中存储的大量图片处理方法，可以在已有的图片识别基础上，增加不同质量值的判断，从而实现对垃圾图片识别的准确率。
在阅读并理解了附图和详细描述后，可以明白其他方面。
附图概述
图1为本发明实施例图片处理装置示意图；
图2为本发明实施例图片处理装置在不同设备上的进行同步处理结构示意图；
图3为本发明实施例图片处理方法流程示意图。
本发明的实施方式
下文中将结合附图对本发明的实施例进行详细说明。需要说明的是，在不冲突的情况下，本申请中的实施例及实施例中的特征可以相互任意组合。
在附图的流程图示出的步骤可以在诸如一组计算机可执行指令的计算机系统中执行。并且，虽然在流程图中示出了逻辑顺序，但是在某些情况下，可以以不同于此处的顺序执行所示出或描述的步骤。
图1为本发明实施例图片处理装置示意图，如图1所示，包括差质图片检测模块11、本地程序监控模块12、及统计中心模块13。
差质图片检测模块12，设置为：识别存储的图片的瑕疵特征，对所述图片的瑕疵特征进行内容质量值评估获取图片的内容质量值并将所述图片的内容质量值发送至统计中心模块13。
本地程序监控模块12，设置为：记录所述图片在应用程序中的使用参数值，根据使用参数值对所述图片进行第一社交质量值评估获取图片的第一社交质量值并将所述图片对应的第一社交质量值发送至统计中心模块13。可选地，该第一社交质量值越大，表示图片的质量越高，因此该第一社交质量值为正值。
统计中心模块13，设置为：接收所述图片的内容质量值和第一社交质量值，根据所述图片的内容质量值和第一社交质量值计算所述图片的图片质量。
可选地，图片在应用程序中的使用参数值包括如下至少之一：在应用程 序中被访问的次数和访问的时间长度，其中，应用程序可以为拍照程序，图片浏览程序，微博，微信，QQ、图片编辑等等。瑕疵图片特征包括至少其中之一：图片虚焦特征和图片闭眼特征；其他瑕疵特征也可以是红眼、曝光等特征，不仅仅限制于上述瑕疵特征；其中，虚焦特征可以基于差分算子法，最佳曲线拟合，prewitt算子法等算法计算得出，闭眼特征可以基于纹理分割的方法和adaboost算法得出。本发明实施例图片处理装置可以运用在手机、平板、个人电脑等终端设备，但不限于上述设备。
统计中心模块13可包括权重管理子模块131和综合计算子模块132；权重管理子模块131，设置为：对所述图片的内容质量值和第一社交质量值分别分配权重；综合计算子模块132，设置为：根据图片的内容质量值及对应的权重和第一社交质量值及对应的权重计算图片的图片质量值。还可设置为：当本地设备具有重复存储的图片时，对重复存储的图片质量值进行增加。
统计中心模块13还可包括联网通讯子模块133及图片重复检测子模块134；联网通讯子模块133，设置为：识别与本地设备同步的其他设备，接收其他设备发送的设备标识及对应的图片质量值；图片重复检测子模块134，设置为：检测本地设备上是否有与其他设备重复存储的图片。综合计算子模块132，还设置为：当本地设备具有重复存储的图片时，对重复存储的图片质量值进行更新，本实施例采用增加质量值。
所述联网通讯子模块133，还可设置为：当本地设备具有重复存储的图片时，通知具有重复存储的图片的其他设备。
如图2所示，为本发明实施例图片处理装置在不同设备上的进行同步处理结构示意图。
可以采用选举的方式选举其中处理能力较强的终端作图片质量值统筹，例如电脑，电脑接收其他设备已经计算好的图片质量值，按照一定的算法计算图片的唯一特征值，类似图片DNA，尽量唯一，同时记录于电脑图片的记录中，并对重复存储的图片质量值进行增加。其他设备可以为手机、电脑、平板等，在接收到电脑发送的重复存储的图片时，也响应重新计算每种设备上的图片的图片质量。
统计中心模块13还可包括删除子模块135，包括标记单元，设置为：根据预设的图片质量阈值，将低于所述图片质量阈值的图片标记为待删除图片；删除单元，设置为：当接收到删除命令时，删除所述待删除图片。
图片处理装置还可包括社交关系监控模块14，设置为：根据联系人的通讯参数值，对所述图片进行第二社交质量评估获取第二社交质量值。可选地，该第二社交质量值值越大，表示图片的质量越高，因此该第二社交质量值值为正值。
其中，社交关系监控模块14包括通讯录子模块141和人脸识别模块142，通讯录子模块141设置为：记录联系人的通讯参数值，人脸识别模块142设置为：检测图片中是否存在通讯记录中的联系人图片；当检测到存储的图片中存在通讯记录中的联系人的图片时，根据记录联系人的通讯参数值对所述图片进行第二社交质量值评估获取第二社交质量值，并将所述图片的第二社交质量值发送至统计中心模块13；统计中心模块13，还设置为：接收社交关系监控模块14发送的第二社交质量值，其中，权重管理子模块131还设置为：为存储图片的第二社交质量值分配权重；综合计算子模块132，还设置为：根据内容质量值及对应的权重、第一社交质量值及对应的权重、及第二社交质量值及对应的权重计算存储图片的图片质量。通讯参数值可包括如下至少之一：本周通讯次数、本月通讯次数、本年通讯次数之和、及通讯总时长。
表1
Figure PCTCN2016091060-appb-000001
Figure PCTCN2016091060-appb-000002
表1为本发明实施例图片处理装置的算法示意图。计算方式如下：
差质图片检测模块的结果可信度80*总权重0.2+图片浏览程序/使用次数15*总权重0.4*分权重0.33*分权重0.5+图片浏览程序/使用时间100*总权重0.4*分权重0.33*分权重0.5+微博/使用次数13*总权重0.4*分权重0.33*分权重0.5+微博/使用时间25*总权重0.4*分权重0.33*分权重0.5+微信/使用次数25*总权重0.4*分权重0.33*分权重0.5+微博/使用时间15*总权重0.4*分权重0.33*分权重0.5+社交关系监控模块/通话次数10*0.2+重复监控/重复数5*0.2＝31.818；
本发明实施例图片处理装置可增加对标识出来的图片，具有人工手动检查的功能，如果发现本图片处理装置检测的图片不准确，可以进行权重调整，以便得到更精确的过滤结果。上述图片的质量值及权重可以为默认值，也可为用户手动设置。
图3为本发明实施例图片处理方法流程示意图，如图3所示，所述方法包括：
S101：识别存储的图片的瑕疵特征，对所述图片的瑕疵特征进行内容质量值评估获取图片的内容质量值；
S102：记录所述图片在应用程序中的使用参数值，根据所述使用参数值对所述图片进行第一社交质量值评估获取图片的第一社交质量值；可选地，该第一社交质量值越大，表示图片的质量越高，因此该第一社交质量值为正值。
S103：根据所述图片的内容质量值和第一社交质量值计算存储图片的图片质量。其中，包括对所述图片的内容质量值和第一社交质量值分别分配权重；根据存储图片的内容质量值及对应的权重和第一社交质量值及对应的权 重计算所述图片的图片质量值。
还可包括：S104:根据存储图片的图片质量值，确定待删除的图片，其中，包括根据预设的图片质量阈值，将低于所述图片质量阈值的图片标记为待删除图片；当接收到删除命令时，删除所述待删除图片。
可选地，图片在应用程序中的使用参数值包括如下至少之一：在应用程序中被访问的次数和访问的时间长度，其中，应用程序可以为拍照程序，图片浏览程序，微博，微信，QQ、图片编辑等等。瑕疵图片特征包括至少其中之一：图片虚焦特征和图片闭眼特征；其他瑕疵特征也可以是红眼、曝光等特征，不仅仅限制于上述瑕疵特征；其中，虚焦特征可以基于差分算子法，最佳曲线拟合，prewitt算子法等算法计算得出，闭眼特征可以基于纹理分割的方法和adaboost算法得出。本发明实施例图片处理方法可以运用在手机、平板、个人电脑等终端设备，但不限于上述设备。
可选的，上述图片处理方法还包括：根据联系人的通讯参数值，对所述图片进行第二社交质量评估；对第二社交质量值分配权重，根据图片的内容质量值、第一社交质量值、及第二社交质量值计算存储图片的图片质量。可选地，第二社交质量值越大，表示图片的质量越高，因此该值为正值。
其中，根据联系人的通讯参数值，对所述图片进行第二社交质量评估获取第二社交质量值，可包括：监控联系人的通讯记录参数值，并且检测图片中是否存在通讯记录中的联系人图片；当检测到存储的图片中存在通讯记录中的联系人的图片时，根据所述联系人的通讯记录参数值对所述图片进行第二社交质量值评估获取第二社交质量值。
可选的，上述图片处理方法还包括：识别同步的其他设备，接收其他设备发送设备标识及对应的图片质量值；检测本地设备上是否有与其他设备重复存储的图片；当本地设备具有重复存储的图片时，对重复图片的图片质量值进行更新，和当本地设备具有重复存储的图片时，通知具有重复存储的图片的其他设备更新重复图片的图片质量值；
可以采用选举的方式选举其中处理能力较强的终端作图片质量值统筹，例如电脑，电脑接收其他设备已经计算好的图片质量值，按照一定的算法计算图片的唯一特征值，例如：像素均值哈希算法，首先将图片，切割并缩小 到标准的4个8*8图片，将每块图片的像素的与像素均值比较，每个像素如果大于均值该像素基设为1，小于设为0，则一个8*8的图片则有64个1或0的字符串，来表示该图片块DNA，一个整体的图片由256个字符串表示该图片的特征值。类似图片DNA，尽量唯一，同时记录于电脑图片的记录中，并对重复存储的图片质量值进行增加。其他设备在接收到电脑发送的重复存储的图片时，也响应重新计算每种设备上的图片的图片质量。
本发明实施例图片处理方法可增加对标识出来的图片，具有人工手动检查的功能，如果发现本图片处理方法检测的图片不准确，可以进行权重调整，以便得到更精确的过滤结果。
本发明实施例还提供一种计算机可读存储介质，存储有计算机可执行指令，所述计算机可执行指令被处理器执行时实现上述图片处理方法。
本领域普通技术人员可以理解上述实施例的全部或部分步骤可以使用计算机程序流程来实现，所述计算机程序可以存储于一计算机可读存储介质中，所述计算机程序在相应的硬件平台上(如系统、设备、装置、器件等)执行，在执行时，包括方法实施例的步骤之一或其组合。
可选地，上述实施例的全部或部分步骤也可以使用集成电路来实现，这些步骤可以被分别制作成一个个集成电路模块，或者将它们中的多个模块或步骤制作成单个集成电路模块来实现。
上述实施例中的装置/功能模块/功能单元可以采用通用的计算装置来实现，它们可以集中在单个的计算装置上，也可以分布在多个计算装置所组成的网络上。
上述实施例中的装置/功能模块/功能单元以软件功能模块的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。上述提到的计算机可读取存储介质可以是只读存储器，磁盘或光盘等。
虽然本申请所揭露的实施方式如上，但所述的内容仅为便于理解本申请而采用的实施方式，并非用以限定本申请。任何本申请所属领域内的技术人员，在不脱离本申请所揭露的精神和范围的前提下，可以在实施的形式及细 节上进行任何的修改与变化，但本申请的专利保护范围，仍须以所附的权利要求书所界定的范围为准。
工业实用性
本发明实施例通过采集用户对图片的操作行为，作为权重指标提出一种主要针对设备中存储的大量图片处理方法，可以在已有的图片识别基础上，增加不同质量值的判断，从而实现对垃圾图片识别的准确率。

Claims (19)

  1. 一种图片处理装置，包括：差质图片检测模块、本地程序监控模块、及统计中心模块；
    差质图片检测模块，设置为：识别存储的图片的瑕疵特征，根据所述图片的瑕疵特征获取所述图片的内容质量值并将所述图片的内容质量值发送至统计中心模块；
    本地程序监控模块，设置为：记录所述图片在应用程序中的使用参数值，根据所述图片的使用参数值获取所述图片的第一社交质量值并将所述图片的第一社交质量值发送至统计中心模块；
    统计中心模块，设置为：接收所述图片的内容质量值和第一社交质量值，根据所述图片的内容质量值和第一社交质量值计算所述图片的图片质量。
  2. 如权利要求1所述的图片处理装置，其中，所述统计中心模块包括权重管理子模块和综合计算子模块；
    权重管理子模块，设置为：对所述图片的内容质量值和第一社交质量值分别分配权重；
    综合计算子模块，设置为：根据图片的内容质量值及对应的权重和第一社交质量值及对应的权重计算图片的图片质量值。
  3. 如权利要求1所述的图片处理装置，其中，所述统计中心模块还包括联网通讯子模块和图片重复检测子模块；
    联网通讯子模块，设置为：识别与本地设备同步的其他设备，接收其他设备发送的设备标识及对应的图片质量值；
    图片重复检测子模块，设置为：检测本地设备上是否有与其他设备重复存储的图片；
    综合计算子模块，还设置为：当本地设备与其他设备具有重复图片时，更新重复图片的图片质量值。
  4. 如权利要求3所述的图片处理装置，其中，所述联网通讯子模块，还 设置为：当本地设备与其他设备具有重复图片时，通知具有重复图片的其他设备更新重复图片的图片质量值。
  5. 如权利要求1所述的图片处理装置，其中，所述统计中心模块还包括删除子模块，设置为：根据所述图片的图片质量值，确定待删除的图片。
  6. 如权利要求5所述的图片处理装置，其中，所述删除子模块包括：
    标记单元，设置为：根据预设的图片质量阈值，将低于所述图片质量阈值的图片标记为待删除图片；
    删除单元，设置为：当接收到删除命令时，删除所述待删除图片。
  7. 如权利要求1所述的图片处理装置，所述装置还包括：
    社交关系监控模块，设置为：根据联系人的通讯参数值，获得第二社交质量值并发送至统计中心模块；
    统计中心模块，还设置为：接收社交关系监控模块发送的第二社交质量值，其中，权重管理子模块还设置为：为存储图片的第二社交质量值分配权重；
    综合计算子模块，还设置为：根据内容质量值及对应的权重、第一社交质量值及对应的权重、及第二社交质量值及对应的权重计算存储图片的图片质量。
  8. 如权利要求7所述的图片处理装置，其中，所述社交关系监控模块包括通讯录子模块和人脸识别模块；
    通讯录子模块，设置为：记录联系人的通讯参数值；
    人脸检测模块，设置为：检测图片中是否存在通讯记录中的联系人图片；当检测到存储的图片中存在通讯记录中的联系人的图片时，根据记录联系人的通讯参数值获取第二社交质量值，并将所述图片的第二社交质量值发送至统计中心模块。
  9. 一种图片处理方法，所述方法包括：
    识别存储的图片的瑕疵特征，根据所述图片的瑕疵特征获取所述图片的内容质量值；
    记录所述图片在应用程序中的使用参数值，根据所述使用参数值获取第一社交质量值；
    根据所述图片的内容质量值和第一社交质量值计算存储图片的图片质量。
  10. 如权利要求9所述的图片处理方法，其中，所述根据图片的内容质量值和第一社交质量值计算存储图片的图片质量，包括：
    对所述图片的内容质量值和第一社交质量值分别分配权重；
    根据存储图片的内容质量值及对应的权重和第一社交质量值及对应的权重计算所述图片的图片质量值。
  11. 如权利要求9所述的图片处理方法，所述方法还包括：
    识别同步的其他设备，接收其他设备发送的设备标识及对应的图片质量值；
    检测本地设备上是否有与其他设备重复存储的图片；
    当本地设备具有与其他设备重复的图片，更新重复的图片的图片质量值。
  12. 如权利要求11所述的图片处理方法，所述方法还包括：
    当本地设备具有重复存储的图片时，通知具有重复图片的其他设备更新重复图片的质量值。
  13. 如权利要求9所述的图片处理方法，所述方法还包括：根据所述图片的图片质量值，确定待删除的图片。
  14. 如权利要求13所述的图片处理方法，其中，所述根据所述图片的图片质量值，确定待删除的图片包括：
    根据预设的图片质量阈值，将低于所述图片质量阈值的图片标记为待删除图片；
    当接收到删除命令时，删除所述待删除图片。
  15. 如权利要求9所述的图片处理方法，所述方法还包括：
    根据联系人的通讯参数值，获取所述图片的第二社交质量；
    对存储图片的第二社交质量值分配权重；
    根据内容质量值及对应的权重、第一社交质量值及对应的权重、及第二社交质量值及对应的权重计算存储图片的图片质量。
  16. 如权利要求15所述的图片处理方法，其中，所述根据联系人的通讯参数值，获取所述图片进行第二社交质量，包括：
    记录联系人的通讯参数值；
    检测图片中是否存在通讯记录中的联系人图片；当检测到存储的图片中存在通讯记录中的联系人的图片时，根据记录联系人的通讯参数值获取所述图片的第二社交质量值。
  17. 如权利要求16所述的图片处理方法，其中，所述通讯参数值包括如下至少之一：本周通讯次数、本月通讯次数、本年通讯次数之和、及通讯总时长。
  18. 如权利要求9至16任意一项所述的图片处理方法，其中，所述图片在应用程序中的使用参数值包括如下至少之一：在应用程序中被访问的次数和访问的时间长度。
  19. 如权利要求9至16任意一项所述的图片处理方法，其中，所述瑕疵图片特征包括至少其中之一：图片虚焦特征和图片闭眼特征。
PCT/CN2016/091060 2015-07-30 2016-07-22 一种图片处理装置及其方法 WO2017016452A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510458852.XA CN106407872A (zh) 2015-07-30 2015-07-30 一种图片处理装置及其方法
CN201510458852.X 2015-07-30

Publications (1)

Publication Number Publication Date
WO2017016452A1 true WO2017016452A1 (zh) 2017-02-02

Family

ID=57884034

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/091060 WO2017016452A1 (zh) 2015-07-30 2016-07-22 一种图片处理装置及其方法

Country Status (2)

Country Link
CN (1) CN106407872A (zh)
WO (1) WO2017016452A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108959505A (zh) * 2018-06-27 2018-12-07 平安科技（深圳）有限公司 一种管理缓存图片的方法及终端设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010231277A (ja) * 2009-03-25 2010-10-14 Fuji Xerox Co Ltd 画像処理装置、記憶制御装置及びプログラム
CN102262727A (zh) * 2011-06-24 2011-11-30 常州锐驰电子科技有限公司 客户采集终端人脸图像质量实时监控方法
CN103092946A (zh) * 2013-01-11 2013-05-08 中兴通讯股份有限公司 终端批量图片选择的方法及系统
CN103716539A (zh) * 2013-12-16 2014-04-09 乐视致新电子科技（天津）有限公司 一种拍照方法和设备
CN104182975A (zh) * 2014-08-11 2014-12-03 福州瑞芯微电子有限公司 自动过滤效果不佳的图片的拍照装置及方法

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB0818089D0 (en) * 2008-10-03 2008-11-05 Eastman Kodak Co Interactive image selection method
CN102137253A (zh) * 2010-01-25 2011-07-27 华为技术有限公司 图片处理的方法、终端及服务器
CN101917352B (zh) * 2010-06-12 2012-07-25 盈世信息科技(北京)有限公司 识别图片垃圾邮件的方法及系统
US20140233811A1 (en) * 2012-05-15 2014-08-21 Google Inc. Summarizing a photo album
CN103580999A (zh) * 2013-11-26 2014-02-12 北京中久联科技有限公司 一种移动互联网社交平台用户信息处理方法及系统
CN103995889B (zh) * 2014-06-03 2017-11-03 广东欧珀移动通信有限公司 图片分类方法及装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010231277A (ja) * 2009-03-25 2010-10-14 Fuji Xerox Co Ltd 画像処理装置、記憶制御装置及びプログラム
CN102262727A (zh) * 2011-06-24 2011-11-30 常州锐驰电子科技有限公司 客户采集终端人脸图像质量实时监控方法
CN103092946A (zh) * 2013-01-11 2013-05-08 中兴通讯股份有限公司 终端批量图片选择的方法及系统
CN103716539A (zh) * 2013-12-16 2014-04-09 乐视致新电子科技（天津）有限公司 一种拍照方法和设备
CN104182975A (zh) * 2014-08-11 2014-12-03 福州瑞芯微电子有限公司 自动过滤效果不佳的图片的拍照装置及方法

Also Published As

Publication number Publication date
CN106407872A (zh) 2017-02-15

Similar Documents

Publication Publication Date Title
US10628680B2 (en) Event-based image classification and scoring
US20230319130A1 (en) Systems and methods for selecting media items
EP2902941B1 (en) System and method for visually distinguishing faces in a digital image
WO2017000716A2 (zh) 图片的管理方法、装置及终端设备
WO2015043512A1 (zh) 一种图片管理方法和装置
WO2017107672A1 (zh) 一种信息处理方法和装置、一种用于信息处理的装置
JPWO2013031096A1 (ja) 画像処理装置、画像処理方法、プログラム、集積回路
CN105472239A (zh) 照片处理方法及装置
CN105243371A (zh) 一种人脸美颜程度的检测方法、系统及拍摄终端
WO2016197657A1 (zh) 一种拍照处理方法和装置
CN105979363A (zh) 一种身份识别法和装置
CN106575223A (zh) 图像分类方法和图像分类装置
US8897484B1 (en) Image theft detector
WO2018028253A1 (zh) 一种信息处理方法、装置及存储介质
WO2023056896A1 (zh) 清晰度的确定方法、装置及设备
WO2015196681A1 (zh) 一种图片处理方法及电子设备
CN109040594A (zh) 拍照方法及装置
WO2017177559A1 (zh) 一种图像管理方法和装置
CN113255685A (zh) 一种图像处理方法、装置、计算机设备以及存储介质
CN109886239B (zh) 人像聚类方法、装置及系统
CN108932703B (zh) 图片处理方法、图片处理装置及终端设备
WO2017016452A1 (zh) 一种图片处理装置及其方法
TWI547816B (zh) 照片分群系統及方法
CN106161743B (zh) 一种媒体资源处理方法、装置及终端
US20170075890A1 (en) Content Ranking Based on Person-to-Person Sharing

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16829818

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16829818

Country of ref document: EP

Kind code of ref document: A1