CN109558779A - Image detecting method and device - Google Patents
Image detecting method and device Download PDFInfo
- Publication number
- CN109558779A CN109558779A CN201810734680.8A CN201810734680A CN109558779A CN 109558779 A CN109558779 A CN 109558779A CN 201810734680 A CN201810734680 A CN 201810734680A CN 109558779 A CN109558779 A CN 109558779A
- Authority
- CN
- China
- Prior art keywords
- image
- human body
- detected
- default human
- default
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the present application discloses image detecting method and device.One specific embodiment of this method includes: acquisition image to be detected；Default human body identification is carried out to image to be detected, obtains the location information of at least one default human body image for including in image to be detected；Based on obtained location information, the default human body image of preset quantity is intercepted from image to be detected；Pre-set categories image classification is carried out to the default human body image and image to be detected of interception respectively, obtains sorting result information；Based on sorting result information, generate for characterize image to be detected whether be pre-set categories image testing result information.The embodiment realizes the accuracy rate for improving testing result information.
Description
Technical field
The invention relates to field of computer technology, and in particular to image detecting method and device.
Background technique
With the fast development of internet, especially mobile Internet is universal, and the video or image layer of various contents go out
It is not poor.Currently, the mode of manual examination and verification is mainly taken to audit the content of these videos or image.
Summary of the invention
The embodiment of the present application proposes image detecting method and device.
In a first aspect, the embodiment of the present application provides a kind of image detecting method, this method comprises: obtaining mapping to be checked
Picture；Default human body identification is carried out to image to be detected, obtains at least one the default human body portion for including in image to be detected
The location information of bit image；Based on obtained location information, the default human body of preset quantity is intercepted from image to be detected
Image；Pre-set categories image classification is carried out to the default human body image and image to be detected of interception respectively, obtains classification knot
Fruit information；Based on sorting result information, generate for characterize image to be detected whether be pre-set categories image testing result letter
Breath.
In some embodiments, default human body identification is carried out to image to be detected, obtain include in image to be detected
At least one default human body image location information, comprising: by image to be detected input training in advance, for identification
The default human body identification model of default human body image, obtains recognition result, recognition result includes in image to be detected
Including at least one default human body image location information.
In some embodiments, recognition result further include: the default human body of at least one for including in image to be detected
The classification information and confidence level of human body shown by image.
In some embodiments, based on obtained location information, from the default human body of image to be detected interception preset quantity
Position image, comprising: human body image is preset at least one, according to the sequence that confidence level is descending, based on obtaining
Location information, from image to be detected intercept preset quantity default human body image.
In some embodiments, pre-set categories figure is carried out to the default human body image and image to be detected of interception respectively
As classification, sorting result information is obtained, comprising: respectively input the default human body image and image to be detected of interception preparatory
It is trained, for determine image whether be pre-set categories image pre-set categories image classification model, obtain sorting result information.
Second aspect, the embodiment of the present application provide a kind of image detection device, which includes: acquiring unit, are matched
It is set to acquisition image to be detected；Recognition unit is configured to carry out image to be detected default human body identification, obtains to be checked
The location information of the default human body image of at least one for including in altimetric image；Interception unit is configured to based on obtaining
Location information intercepts the default human body image of preset quantity from image to be detected；Taxon, it is right respectively to be configured to
The default human body image and image to be detected of interception carry out pre-set categories image classification, obtain sorting result information；It generates
Unit, be configured to generate based on sorting result information for characterize image to be detected whether be pre-set categories image detection
Result information.
In some embodiments, recognition unit is further configured to: by image to be detected input in advance training, be used for
The default human body identification model for identifying default human body image, obtains recognition result, recognition result includes mapping to be checked
The location information of the default human body image of at least one for including as in.
In some embodiments, recognition result further include: the default human body of at least one for including in image to be detected
The classification information and confidence level of human body shown by image.
In some embodiments, interception unit is further configured to: being preset human body image at least one, is pressed
According to the sequence that confidence level is descending, based on obtained location information, the default people of preset quantity is intercepted from image to be detected
Body region image.
In some embodiments, taxon is further configured to: respectively by the default human body image of interception and
Image to be detected input in advance training, for determine image whether be pre-set categories image pre-set categories image classification mould
Type obtains sorting result information.
The third aspect, the embodiment of the present application provide a kind of electronic equipment, which includes: one or more processing
Device；Storage device is stored thereon with one or more programs；When said one or multiple programs are by said one or multiple processing
Device executes, so that said one or multiple processors realize the method as described in implementation any in first aspect.
Fourth aspect, the embodiment of the present application provide a kind of computer-readable medium, are stored thereon with computer program, on
State the method realized as described in implementation any in first aspect when program is executed by processor.
Image detecting method and device provided by the embodiments of the present application carry out default human body to image to be detected first
Identification, obtains the location information of default human body image for including in image to be detected.So as to based on obtained position
Information intercepts the default human body image of preset quantity from image to be detected.Later respectively to the default human body portion of interception
Bit image and image to be detected carry out pre-set categories image classification, obtain sorting result information.It is finally based on sorting result information,
Generate for characterize image to be detected whether be pre-set categories image testing result information.Since testing result information is to be based on
What image to be detected and the default human body image of interception generated.Therefore, testing result informix image to be detected
Global Information and local message, to improve the accuracy rate of testing result information.
Detailed description of the invention
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the application's is other
Feature, objects and advantages will become more apparent upon:
Fig. 1 is that one embodiment of the application can be applied to exemplary system architecture figure therein；
Fig. 2 is the flow chart according to one embodiment of the image detecting method of the application；
Fig. 3 is the schematic diagram according to an application scenarios of the image detecting method of the application；
Fig. 4 is the flow chart according to another embodiment of the image detecting method of the application；
Fig. 5 is the structural schematic diagram according to one embodiment of the image detection device of the application；
Fig. 6 is adapted for the structural schematic diagram for the computer system for realizing the electronic equipment of the embodiment of the present application.
Specific embodiment
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, part relevant to related invention is illustrated only in attached drawing.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 is shown can be using the image detecting method of the embodiment of the present application or the exemplary system of image detection device
Framework 100.
As shown in Figure 1, system architecture 100 may include terminal 101,102,103, network 104 and server 105.Network
104 between terminal device 101,102,103 and server 105 to provide the medium of communication link.Network 104 can wrap
Include various connection types, such as wired, wireless communication link or fiber optic cables etc..
Terminal device 101,102,103 is interacted by network 104 with server 105, such as image to be detected is sent to
Server 105.All kinds of applications of taking pictures, picture processing application etc. can be installed on terminal device 101,102,103.
Terminal device 101,102,103 can be hardware, be also possible to software.When terminal device 101,102,103 is hard
When part, can be the equipment that can shoot or store image, including but not limited to: camera, the mobile phone for having camera function,
Picture storage server etc..When terminal device 101,102,103 is software, it may be mounted at above-mentioned cited electronics and set
In standby.Multiple softwares or software module (such as providing the service of taking pictures) may be implemented into it, also may be implemented into single soft
Part or software module.It is not specifically limited herein.
Server 105 can be to provide the server of various services, such as obtain to from terminal device 101,102,103
Image to be detected is detected, and testing result information is obtained.It is set if desired, result information can also be will test and be sent to terminal
Standby 101,102,103.
It should be noted that image detecting method provided by the embodiment of the present application can be executed by server 105, it can also
To be executed by terminal device.Correspondingly, image detection device can be set in server 105, also can be set and sets in terminal
In standby.
It should be noted that can also be detected to image in terminal device 101,102,103.At this point, image detection
Method can also be executed by terminal device 101,102,103.Correspondingly, image detection device also can be set in terminal device
101, in 102,103.At this point, server 105 and network 104 can be not present in exemplary system architecture 100.
It should be noted that server can be hardware, it is also possible to software.When server is hardware, may be implemented
At the distributed server cluster that multiple servers form, individual server also may be implemented into.It, can when server is software
To be implemented as multiple softwares or software module (such as providing Distributed Services), single software or software also may be implemented into
Module.It is not specifically limited herein.
It should be understood that the number of terminal device, network and server in Fig. 1 is only schematical.According to realization need
It wants, can have any number of terminal device, network and server.
With continued reference to Fig. 2, the process 200 of one embodiment of the image detecting method according to the application is shown.The figure
As detection method, comprising the following steps:
Step 201, image to be detected is obtained.
In the present embodiment, the executing subject of image detecting method can pass through wired connection mode or wireless connection side
Formula obtains image to be detected from terminal device.Wherein, image to be detected can be arbitrary image.The determination of image to be detected can be with
It is specified by technical staff, it can also be according to certain conditional filtering.In addition, image to be detected may be stored in above-mentioned execution master
Body is local.At this point, above-mentioned executing subject directly can locally obtain image to be detected.
Step 202, default human body identification is carried out to image to be detected, obtains include in image to be detected at least one
The location information of a default human body image.
In the present embodiment, above-mentioned executing subject can carry out default human body to image to be detected by various methods
Identification obtains the location information of at least one default human body image for including in image to be detected.Wherein, human body portion is preset
Position can be at least one position of human body, including but not limited at least one of following: mouth, eyes, nose etc..Wherein,
Location information is for characterizing position of the default human body image relative to image to be detected.Location information has various forms, example
Such as, callout box, coordinate etc..
As an example, above-mentioned executing subject can carry out default human body identification by cascade classifier, to obtain
The location information of the default human body image of at least one for including in image to be detected.Wherein, cascade classifier can be grade
Multiple classifiers (such as Haar classifier) of connection.As an example, OpenCV (Open Source Computer Vision
Library, the cross-platform computer vision library of open source) in include many different parts for human body (for example, lip, eyes
Deng) classifier, can according to need selection.The verification and measurement ratio of multiple classifiers is different.Verification and measurement ratio, which can be, to be detected in image
The probability of the default human body of display.Multiple classifiers can be cascaded according to the descending sequence of verification and measurement ratio.Specifically
For, image to be detected can be inputted cascade classifier by above-mentioned executing subject.It is treated first by the maximum classifier of verification and measurement ratio
Detection image is detected, if detecting, image to be detected shows default human body, and image to be detected is sent to next fraction
Class device.Subsequent classifier is detected in the same way, until the smallest classifier of verification and measurement ratio.If detecting figure
Without showing default human body as in, it can characterize to obtain the information that default human body is not shown in image.If last
First-level class device (the smallest classifier of verification and measurement ratio), which detects, shows default human body in image, can export to indicate to scheme
The callout box of the location information of default human body shown in as in.At least one to obtain including in image to be detected is pre-
If the location information of human body image.
In some optional implementations of the present embodiment, can by image to be detected input in advance training, be used for
The default human body identification model for identifying default human body image, obtains recognition result, recognition result includes mapping to be checked
The location information of the default human body image of at least one for including as in.Wherein, presetting human body identification model can use
Whether comprising default human body image in detection image, i.e., default human body whether is shown in image.
In these implementations, default human body identification model can be obtained by following steps training:
The first step, the available training sample set of executing subject.Wherein, each training sample in training sample set
It may include the markup information of sample image and sample image.Markup information includes the default human body for including in sample image
Image location information and sample image in the classification information of default human body that shows.Wherein, location information is used for table
Levy position of the image of default human body relative to sample image.Location information has various forms, for example, callout box, coordinate
Etc..The classification information of default human body is used to indicate the classification of default human body.Wherein, default human body can be
At least one human body.As an example, default human body may include mouth, eyes, nose.So, human body portion is preset
The classification information of position can be " 00 ", " 01 ", " 10 ", be respectively used to indicate these three positions.
Second step, executing subject can be using the sample images of the training sample in training sample set as input, will be with
For the corresponding markup information of the sample image of input as desired output, training obtains default human body detection model.It is specific next
It says, the sample image of the training sample in training sample set can be inputted into initial preset human body detection model.Wherein,
Initial preset human body detection model can be various target detection networks.As an example, can be existing SSD
(Single Shot MultiBox Detector) or YOLO (You Only Look Once) etc..In practice, Ke Yiwei
Initial value is arranged in initial preset human body detection model.For example, it may be some different small random numbers." small random number " is used
Guarantee that network will not enter saturation state because weight is excessive, so as to cause failure to train, " difference " is used to guarantee that network can
Normally to learn.Later, the testing result of the sample image of available input.Believed with the mark of the sample image with input
The desired output as initial preset human body detection model is ceased, machine learning method training initial preset human body is utilized
Detection model.It specifically, can be first between the preset loss function testing result being calculated and markup information
Difference.It is then possible to be based on obtained difference, the parameter of initial preset human body detection model is adjusted, and is being met
In the case where preset trained termination condition, terminate training, and using the initial preset human body detection model after training as
Default human body detection model.Here training termination condition includes but is not limited at least one of following: the training time is more than
Preset duration；Frequency of training reaches preset times；It calculates resulting difference and is less than default discrepancy threshold.
Here it can adopt in various manners based on obtained testing result mark letter corresponding with the training sample of input
Difference between breath adjusts the parameter of initial preset human body detection model.For example, BP (Back can be used
Propagation, backpropagation) algorithm or SGD (Stochastic Gradient Descent, stochastic gradient descent) calculate
Method adjusts the parameter of initial pictures sorter network.
It should be noted that the executing subject of training step can be identical with the executing subject of image detecting method, it can also
With difference.If they are the same, executing subject can detect default human body after training obtains default human body detection model
The network structure and parameter value of model are stored in local.If it is different, the executing subject of training step obtains default human body in training
After location detection model, the network structure of model and parameter value can be sent to the executing subject of image detecting method.
Step 203, based on obtained location information, the default human body figure of preset quantity is intercepted from image to be detected
Picture.
In the present embodiment, above-mentioned executing subject is based on obtained location information, using various methods (for example, various sections
Figure application) the default human body image of interception preset quantity from image to be detected.As an example, location information can be square
Shape callout box.In practice, rectangle callout box be can be represented by vectors.It may include the geometric center of rectangle callout box in vector
Coordinate, the height of rectangle callout box and width.
Step 204, pre-set categories image classification is carried out to the default human body image and image to be detected of interception respectively,
Obtain sorting result information.
In the present embodiment, above-mentioned executing subject can use various methods respectively to the default human body image of interception
Pre-set categories image classification is carried out with image to be detected, obtains sorting result information.Wherein, pre-set categories image can be various
The image of classification.For example, it may be face-image, bad image etc..It should be noted that pre-set categories image herein with
Default human body image matches.Specifically, if pre-set categories image is face-image.So, human body portion is preset
Bit image can be mouth image, eye image etc..Wherein, sorting result information seems no for pre-set categories for phenogram
Image.Sorting result information can be various forms of information.As an example, sorting result information can be numerical value.For example, with
" 0 " indicates not being pre-set categories image, indicates to be pre-set categories image with " 1 ".As an example, sorting result information can also be
Text, character etc..
As an example, above-mentioned executing subject can use disaggregated model (such as word packet model), interception is preset respectively
Human body image and image to be detected carry out pre-set categories image classification, obtain sorting result information.Wherein, word packet model is answered
It is widely used in image recognition, realization be may include feature extraction, feature coding, feature convergence and be divided using classifier
Class.Specifically, feature extraction can use various detective operators, such as Harris corner detection operator, FAST
(Features from Accelerated Segment Test) operator etc. carries out feature extraction to object.On this basis,
In order to improve the robustness of feature representation, feature can be encoded.For example, realizing feature coding by inquiry dictionary.
Later, multiple features by coding are spliced, and the final expression as image, specific manifestation form can be
Vector.Finally, being classified using the vector that the classifiers such as support vector machines obtain splicing.To obtain classification results letter
Breath.
Step 205, be based on sorting result information, generate for characterize image to be detected whether be pre-set categories image inspection
Survey result information.
In the present embodiment, above-mentioned executing subject can be based on sorting result information using a variety of methods, generate and be used for table
Sign image to be detected whether be pre-set categories image testing result information.
As an example, obtaining sorting result information in step 204 may include the first sorting result information and the second classification
Result information.Wherein, whether the default human body image that the first sorting result information is used to characterize interception is pre-set categories figure
Picture.Second sorting result information is for characterizing whether image to be detected is pre-set categories image.Above-mentioned executing subject can be by
One sorting result information and the second sorting result information input classifier.Wherein, as an example, classifier can be Softmax
Classifier.Softmax can will be inputted on information MAP to (0,1) section.And it can be used in calculating process all defeated
Enter information.It later, can be using the output of the corresponding classifier of the second sorting result information as testing result information.
As an example, above-mentioned executing subject can also be by inquiring preset, sorting result information and testing result information
Mapping table, and obtain testing result information.Wherein, mapping table, which can be, is obtained based on a large amount of statistics.It is right
It should be related to that characterization can record a large amount of result information and corresponding testing result information.Above-mentioned executing subject can be for step
Sorting result information obtained in rapid 204, is inquired in mapping table, if it exists with obtained sorting result information
The sorting result information matched, the then corresponding testing result information of the sorting result information that may be matched.Later, it can incite somebody to action
To testing result information as characterize image to be detected whether be pre-set categories image testing result information.
With continued reference to the schematic diagram that Fig. 3, Fig. 3 are according to the application scenarios of the image detecting method of the present embodiment.?
In the application scenarios of Fig. 3, the executing subject of image detecting method can be server 300.Server 300 obtains to be detected first
Image 301.Later, default human body identification is carried out to image to be detected.Obtain the default people for including in image to be detected 301
The location information of the image of body region is shown as callout box 302 in figure.Later, server 300 is cut from image to be detected 301
The image for taking default human body obtains truncated picture 303.Then respectively by truncated picture 303 and image to be detected 301
Disaggregated model (such as word packet model) 304 is inputted, sorting result information is obtained.As shown, truncated picture 303 and to be detected
The sorting result information of image 301 is " 1 ".It indicates truncated picture 303 and image to be detected 301 is pre-set categories image
(such as bad image).On this basis, server 300 can be by inquiring mapping table 305.Pass through inquiry, classification knot
Fruit information is that " 1 " corresponding testing result information is also " 1 ".So as to obtain for characterizing whether image to be detected is pre-
If the testing result information of classification image (such as bad image).
The method provided by the above embodiment of the application can obtain image to be detected first.Later, to image to be detected
Default human body identification is carried out, the position letter of at least one default human body image for including in image to be detected is obtained
Breath.So as to intercept the default human body image of preset quantity from image to be detected based on obtained location information.So
Pre-set categories image classification is carried out to the default human body image and image to be detected of interception respectively afterwards, obtains classification results letter
Breath.Finally, be based on sorting result information, generate for characterize image to be detected whether be pre-set categories image testing result letter
Breath.Since testing result information is generated based on image to be detected and the default human body image of interception.Therefore, detection knot
The fruit informix Global Information and local message of image to be detected, to improve the accuracy rate of testing result information.
With further reference to Fig. 4, it illustrates the processes 400 of another embodiment of image detecting method.The image detection side
The process 400 of method, comprising the following steps:
Step 401, image to be detected is obtained.
In the present embodiment, the step of the specific implementation of step 401 and brought technical effect corresponding with Fig. 2 embodiment
201 is similar, and details are not described herein.
Step 402, default human body image to be detected input being trained in advance, presetting human body image for identification
Position identification model, obtains recognition result.
In the present embodiment, recognition result includes at least one the default human body image for including in image to be detected
The classification information of human body shown by the default human body image of at least one for including in location information, image to be detected
And confidence level.Wherein, confidence level is used to indicate the credibility of classification information.In practice, confidence level can use probability value table
Show.
In the present embodiment, image to be detected can be inputted training in advance, use by the executing subject of image detecting method
In the default human body identification model for identifying default human body image.Wherein, human body identification model is preset for knowing
Default human body image in other image.
Step 403, the sequence descending according to confidence level is cut from image to be detected based on obtained location information
Take the default human body image of preset quantity.
It in the present embodiment, may include at least one default human body image in image to be detected.Therefore, above-mentioned to hold
Row main body can the sequence descending according to confidence level present count is intercepted from image to be detected based on obtained location information
The default human body image of amount.
Step 404, respectively by the default human body image and image to be detected of interception input training in advance, be used for really
Determine image whether be pre-set categories image pre-set categories image classification model, obtain sorting result information.
In the present embodiment, above-mentioned executing subject can be respectively by the default human body image and image to be detected of interception
Input pre-set categories image classification model trained in advance, obtains sorting result information.Wherein, sorting result information can be use
In phenogram seem the no information for pre-set categories image.It should be noted that sorting result information here may include
One sorting result information and the second sorting result information.Wherein, the first sorting result information can be used for characterizing the default of interception
Whether human body image is pre-set categories image.Second sorting result information can be used for characterizing whether image to be detected is pre-
If classification image.
Pre-set categories image can be the image of any classification.As an example, can be facial image, head image, no
Plan deliberately picture etc..Pre-set categories image classification model is for determining whether image is pre-set categories image.
As an example, pre-set categories image classification model can be obtained by following steps training:
The first step obtains training sample set.Wherein, each training sample in training sample set includes sample image
And markup information.Markup information is for characterizing whether sample image is pre-set categories image.Here, markup information can be various
Form.As an example, markup information can be numerical value.For example, indicating not being pre-set categories image with " 0 ", indicate to be pre- with " 1 "
If classification image.As an example, markup information can also be text, character etc..
Second step, will using the sample image of the training sample as input for the training sample in training sample set
Markup information corresponding with the sample image of input is as desired output, and using the method for machine learning, training obtains default class
Other disaggregated model.Specifically, initial preset classification image classification model can be trained based on training sample set, is obtained
To pre-set categories image classification model.Wherein, initial preset classification image classification model can be various image classification networks.Make
For example, residual error network (Deep Residual Network, ResNet), VGG etc. can be.VGG is the view of certain university
Feel the disaggregated model that geometry group (Visual Geometry Group, VGG) is proposed.
Specifically, the sample image of training sample can be inputted into initial preset category classification model.It, can be in practice
For initial preset category classification model, initial value is set.Later, the sorting result information of the sample image of available input.So
It can use the difference between the sorting result information and markup information that preset loss function is calculated afterwards.It later, can be with
Based on obtained difference, the parameter of initial preset category classification model is adjusted, and in the feelings for meeting preset trained termination condition
Under condition, terminate training, using the initial preset category classification model after training as pre-set categories image classification model.Here instruction
It includes but is not limited at least one of following for practicing termination condition: the training time is more than preset duration；Frequency of training reaches preset times；
It calculates resulting difference and is less than default discrepancy threshold.
Step 405, be based on sorting result information, generate for characterize image to be detected whether be pre-set categories image inspection
Survey result information.
In the present embodiment, in the specific implementation of step 405 and brought technical effect embodiment corresponding with Fig. 2
Step 205 it is similar, details are not described herein.
Figure 4, it is seen that compared with the corresponding embodiment of Fig. 2, the image detecting method in the present embodiment pass through by
Image to be detected inputs default human body identification model, obtains recognition result.And the sequence descending according to confidence level, it cuts
Take the default human body image of preset quantity.Compared with other recognition methods, the accurate of default human body identification is improved
Property.The Detection accuracy of image to be detected is improved as a result,.
With further reference to Fig. 5, as the realization to method shown in above-mentioned each figure, this application provides a kind of image detection dresses
The one embodiment set, the Installation practice is corresponding with embodiment of the method shown in Fig. 2, which specifically can be applied to respectively
In kind electronic equipment.
As shown in figure 5, the image detection device 500 of the present embodiment includes: acquiring unit 501, recognition unit 502, interception
Unit 503, taxon 504 and generation unit 505.Wherein, acquiring unit 501 is configured to obtain image to be detected.Identification
Unit 502 is configured to carry out image to be detected default human body identification, obtains include in image to be detected at least one
The location information of a default human body image.Interception unit 503 is configured to based on obtained location information, from mapping to be checked
The default human body image of preset quantity is intercepted as in.Taxon 504 is configured to the default human body portion to interception respectively
Bit image and image to be detected carry out pre-set categories image classification, obtain sorting result information.Generation unit 505 is configured to base
In sorting result information, generate for characterize image to be detected whether be pre-set categories image testing result information.
Acquiring unit 501 that image detection device 600 in the present embodiment includes, recognition unit 502, interception unit
503, the step of the specific implementation of taxon 504 and generation unit 505 and brought technical effect embodiment corresponding with Fig. 2
Rapid 201-205, details are not described herein.
In some optional implementations of the present embodiment, recognition unit is further configured to: by image to be detected
Default human body identification model that input is trained in advance, presetting human body image for identification, obtains recognition result, knows
Other result includes the location information at least one the default human body image for including in image to be detected.
In some optional implementations of the present embodiment, recognition result, which can also include: in image to be detected, includes
At least one default human body image shown by human body classification information and confidence level.
In some optional implementations of the present embodiment, interception unit 503 can be further configured to: for extremely
A few default human body image, according to the sequence that confidence level is descending, based on obtained location information, from mapping to be checked
The default human body image of preset quantity is intercepted as in.
In some optional implementations of the present embodiment, taxon 504 can be further configured to: respectively will
Interception default human body image and image to be detected input in advance training, for determining whether image is pre-set categories figure
The pre-set categories image classification model of picture, obtains sorting result information.
In the present embodiment, above-mentioned image detection device can obtain image to be detected by acquiring unit 501 first.It
Afterwards, recognition unit 502 carries out default human body identification to image to be detected, obtains at least one for including in image to be detected
The location information of default human body image.To which interception unit 503 can be based on obtained location information, from image to be detected
The default human body image of middle interception preset quantity.Then taxon 504 is respectively to the default human body image of interception
Pre-set categories image classification is carried out with image to be detected, obtains sorting result information.Finally, generation unit 505 is based on classification knot
Fruit information, generate for characterize image to be detected whether be pre-set categories image testing result information.Since testing result is believed
Breath is generated based on image to be detected and the default human body image of interception.Therefore, testing result informix is to be checked
The Global Information and local message of altimetric image, to improve the accuracy rate of testing result information.
With continued reference to Fig. 6, it illustrates the computer systems 600 for the electronic equipment for being suitable for being used to realize the embodiment of the present application
Structural schematic diagram.Electronic equipment shown in Fig. 6 is only an example, function to the embodiment of the present application and should not use model
Shroud carrys out any restrictions.
As shown in fig. 6, computer system 600 includes central processing unit (CPU) 601, it can be read-only according to being stored in
Program in memory (ROM) 602 is loaded into the program in random access storage device (RAM) 603 from storage section 608
And execute various movements appropriate and processing.In RAM 603, also it is stored with system 600 and operates required various program sum numbers
According to.CPU 601, ROM 602 and RAM 603 are connected with each other by bus 604.Input/output (I/O) interface 605 also connects
To bus 604.
I/O interface 605 is connected to lower component: the importation 606 including keyboard, mouse etc.；It is penetrated including such as cathode
The output par, c 607 of spool (CRT), liquid crystal display (LCD) etc. and loudspeaker etc.；Storage section 608 including hard disk etc.；
And the communications portion 609 of the network interface card including LAN card, modem etc..Communications portion 609 via such as because
The network of spy's net executes communication process.Driver 610 is also connected to I/O interface 605 as needed.Detachable media 611, it is all
Such as disk, CD, magneto-optic disk, semiconductor memory are mounted on as needed on driver 610, in order to read from thereon
Computer program out is mounted into storage section 608 as needed.
Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description
Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium
On computer program, which includes the program code for method shown in execution flow chart.In such reality
It applies in example, which can be downloaded and installed from network by communications portion 609, and/or from detachable media
611 are mounted.When the computer program is executed by central processing unit (CPU) 601, executes and limited in the present processes
Above-mentioned function.
It should be noted that computer-readable medium described herein can be computer-readable signal media or meter
Calculation machine readable storage medium storing program for executing either the two any combination.Computer readable storage medium for example can be --- but not
Be limited to --- electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor system, device or device, or any above combination.Meter
The more specific example of calculation machine readable storage medium storing program for executing can include but is not limited to: have the electrical connection, just of one or more conducting wires
Taking formula computer disk, hard disk, random access storage device (RAM), read-only memory (ROM), erasable type may be programmed read-only storage
Device (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device,
Or above-mentioned any appropriate combination.In this application, computer readable storage medium can be it is any include or storage journey
The tangible medium of sequence, the program can be commanded execution system, device or device use or in connection.And at this
In application, computer-readable signal media may include in a base band or as carrier wave a part propagate data-signal,
Wherein carry computer-readable program code.The data-signal of this propagation can take various forms, including but unlimited
In electromagnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be that computer can
Any computer-readable medium other than storage medium is read, which can send, propagates or transmit and be used for
By the use of instruction execution system, device or device or program in connection.Include on computer-readable medium
Program code can transmit with any suitable medium, including but not limited to: wireless, electric wire, optical cable, RF etc. are above-mentioned
Any appropriate combination.
The calculating of the operation for executing the application can be write with one or more programming languages or combinations thereof
Machine program code, described program design language include object oriented program language-such as Java, Smalltalk, C+
+, it further include conventional procedural programming language-such as " C " language or similar programming language.Program code can
Fully to execute, partly execute on the user computer on the user computer, be executed as an independent software package,
Part executes on the remote computer or executes on a remote computer or server completely on the user computer for part.
In situations involving remote computers, remote computer can pass through the network of any kind --- including local area network (LAN)
Or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as utilize Internet service
Provider is connected by internet).
Flow chart and block diagram in attached drawing are illustrated according to the system of the various embodiments of the application, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, program segment or code of table, a part of the module, program segment or code include one or more use
The executable instruction of the logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box
The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually
It can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it to infuse
Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding
The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
Being described in unit involved in the embodiment of the present application can be realized by way of software, can also be by hard
The mode of part is realized.Described unit also can be set in the processor, for example, can be described as: a kind of processor packet
Include acquiring unit, recognition unit, interception unit, taxon and generation unit.Wherein, the title of these units is in certain situation
Under do not constitute restriction to the unit itself, for example, acquiring unit is also described as " obtaining the list of image to be detected
Member ".
As on the other hand, present invention also provides a kind of computer-readable medium, which be can be
Included in electronic equipment described in above-described embodiment；It is also possible to individualism, and without in the supplying electronic equipment.
Above-mentioned computer-readable medium carries one or more program, when said one or multiple programs are held by the electronic equipment
When row, so that the electronic equipment: obtaining image to be detected；Default human body identification is carried out to image to be detected, is obtained to be checked
The location information of the default human body image of at least one for including in altimetric image；Based on obtained location information, to be detected
The default human body image of preset quantity is intercepted in image；Respectively to the default human body image and image to be detected of interception
Pre-set categories image classification is carried out, sorting result information is obtained；Based on sorting result information, generate for characterizing image to be detected
Whether be pre-set categories image testing result information.
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.Those skilled in the art
Member is it should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from foregoing invention design, it is carried out by above-mentioned technical characteristic or its equivalent feature
Any combination and the other technical solutions formed.Such as features described above and (but being not limited to) disclosed herein have it is similar
The technical characteristic of function is replaced mutually and the technical solution that is formed.
Claims (12)
1. a kind of image detecting method, comprising:
Obtain image to be detected；
Default human body identification is carried out to described image to be detected, at least one for obtaining including in described image to be detected is pre-
If the location information of human body image；
Based on obtained location information, the default human body image of preset quantity is intercepted from described image to be detected；
Pre-set categories image classification is carried out to the default human body image of interception and described image to be detected respectively, is classified
Result information；
Based on the sorting result information, generate for characterize described image to be detected whether be pre-set categories image detection knot
Fruit information.
2. it is described that default human body identification is carried out to described image to be detected according to the method described in claim 1, wherein,
Obtain the location information of at least one default human body image for including in described image to be detected, comprising:
Default human body identification that the input of described image to be detected is trained in advance, presetting human body image for identification
Model, obtains recognition result, and the recognition result includes at least one the default human body for including in described image to be detected
The location information of image.
3. according to the method described in claim 2, wherein, the recognition result further include:
The classification information of human body shown by the default human body image of at least one for including in described image to be detected
And confidence level.
4. it is described based on obtained location information according to the method described in claim 3, wherein, it is cut from described image to be detected
Take the default human body image of preset quantity, comprising:
Present count is intercepted from described image to be detected based on obtained location information according to the sequence that confidence level is descending
The default human body image of amount.
5. method according to any one of claims 1-4, wherein it is described respectively to the default human body image of interception and
Described image to be detected carries out pre-set categories image classification, obtains sorting result information, comprising:
Respectively by the default human body image of interception and described image to be detected input training in advance, be used to determine image and be
The no pre-set categories image classification model for pre-set categories image, obtains sorting result information.
6. a kind of image detection device, comprising:
Acquiring unit is configured to obtain image to be detected；
Recognition unit is configured to carry out described image to be detected default human body identification, obtains described image to be detected
In include at least one default human body image location information；
Interception unit is configured to based on obtained location information, and the default of preset quantity is intercepted from described image to be detected
Human body image；
Taxon is configured to carry out pre-set categories to the default human body image of interception and described image to be detected respectively
Image classification obtains sorting result information；
Generation unit is configured to generate based on the sorting result information for characterizing whether described image to be detected is pre-
If the testing result information of classification image.
7. device according to claim 6, wherein the recognition unit is further configured to:
Default human body identification that the input of described image to be detected is trained in advance, presetting human body image for identification
Model, obtains recognition result, and the recognition result includes at least one the default human body for including in described image to be detected
The location information of image.
8. device according to claim 7, wherein the recognition result further include:
The classification information of human body shown by the default human body image of at least one for including in described image to be detected
And confidence level.
9. device according to claim 8, wherein the interception unit is further configured to:
For at least one described default human body image, according to the sequence that confidence level is descending, based on obtained position
Information intercepts the default human body image of preset quantity from described image to be detected.
10. according to the device any in claim 6-9, wherein the taxon is further configured to:
Respectively by the default human body image of interception and described image to be detected input training in advance, be used to determine image and be
The no pre-set categories image classification model for pre-set categories image, obtains sorting result information.
11. a kind of electronic equipment, comprising:
One or more processors；
Storage device is stored thereon with one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
Now such as method as claimed in any one of claims 1 to 5.
12. a kind of computer-readable medium, is stored thereon with computer program, wherein real when described program is executed by processor
Now such as method as claimed in any one of claims 1 to 5.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810734680.8A CN109558779A (en) | 2018-07-06 | 2018-07-06 | Image detecting method and device |
| PCT/CN2018/116338 WO2020006964A1 (en) | 2018-07-06 | 2018-11-20 | Image detection method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810734680.8A CN109558779A (en) | 2018-07-06 | 2018-07-06 | Image detecting method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109558779A true CN109558779A (en) | 2019-04-02 |
Family
ID=65864446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810734680.8A Pending CN109558779A (en) | 2018-07-06 | 2018-07-06 | Image detecting method and device |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109558779A (en) |
| WO (1) | WO2020006964A1 (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110310142A (en) * | 2019-05-20 | 2019-10-08 | 浙江新再灵科技股份有限公司 | Elevator card put-on method and device based on crowd's value analysis |
| CN111814852A (en) * | 2020-06-24 | 2020-10-23 | 理光软件研究所（北京）有限公司 | Image detection method, image detection device, electronic equipment and computer-readable storage medium |
| CN113454649A (en) * | 2021-06-17 | 2021-09-28 | 商汤国际私人有限公司 | Target detection method, target detection device, electronic equipment and computer-readable storage medium |
| WO2022263904A1 (en) * | 2021-06-17 | 2022-12-22 | Sensetime International Pte. Ltd. | Target detection methods, apparatuses, electronic devices and computer-readable storage media |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111291640B (en) * | 2020-01-20 | 2023-02-17 | 北京百度网讯科技有限公司 | Method and apparatus for recognizing gait |
| CN112348615A (en) * | 2020-03-24 | 2021-02-09 | 北京沃东天骏信息技术有限公司 | Method and device for auditing information |
| CN111539456B (en) * | 2020-04-02 | 2024-03-01 | 浙江华睿科技股份有限公司 | Target identification method and device |
| CN111935106B (en) * | 2020-07-23 | 2022-05-13 | 湖北工业大学 | Method for collecting labeled data set by using improved verification code |
| CN112580745A (en) * | 2020-12-29 | 2021-03-30 | 北京五八信息技术有限公司 | Image recognition method and device, electronic equipment and computer readable medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101059838A (en) * | 2007-06-11 | 2007-10-24 | 湖北东润科技有限公司 | Human face recognition system and recognition method |
| CN101763502A (en) * | 2008-12-24 | 2010-06-30 | 中国科学院自动化研究所 | High-efficiency method and system for sensitive image detection |
| CN103049733A (en) * | 2011-10-11 | 2013-04-17 | 株式会社理光 | Human face detection method and human face detection equipment |
| CN105844283A (en) * | 2015-01-16 | 2016-08-10 | 阿里巴巴集团控股有限公司 | Method for identifying category of image, image search method and image search device |
| CN107871134A (en) * | 2016-09-23 | 2018-04-03 | 北京眼神科技有限公司 | A kind of method for detecting human face and device |
| CN108229575A (en) * | 2018-01-19 | 2018-06-29 | 百度在线网络技术（北京）有限公司 | For detecting the method and apparatus of target |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102184419B (en) * | 2011-04-13 | 2013-05-01 | 深圳市迈科龙影像技术有限公司 | Pornographic image recognizing method based on sensitive parts detection |
| EP2657857A1 (en) * | 2012-04-27 | 2013-10-30 | ATG Advanced Swiss Technology Group AG | Method for binary classification of a query image |
| CN103366160A (en) * | 2013-06-28 | 2013-10-23 | 西安交通大学 | Objectionable image distinguishing method integrating skin color, face and sensitive position detection |
| CN107992783A (en) * | 2016-10-26 | 2018-05-04 | 上海银晨智能识别科技有限公司 | Face image processing process and device |
-
2018
- 2018-07-06 CN CN201810734680.8A patent/CN109558779A/en active Pending
- 2018-11-20 WO PCT/CN2018/116338 patent/WO2020006964A1/en active Application Filing
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101059838A (en) * | 2007-06-11 | 2007-10-24 | 湖北东润科技有限公司 | Human face recognition system and recognition method |
| CN101763502A (en) * | 2008-12-24 | 2010-06-30 | 中国科学院自动化研究所 | High-efficiency method and system for sensitive image detection |
| CN103049733A (en) * | 2011-10-11 | 2013-04-17 | 株式会社理光 | Human face detection method and human face detection equipment |
| CN105844283A (en) * | 2015-01-16 | 2016-08-10 | 阿里巴巴集团控股有限公司 | Method for identifying category of image, image search method and image search device |
| CN107871134A (en) * | 2016-09-23 | 2018-04-03 | 北京眼神科技有限公司 | A kind of method for detecting human face and device |
| CN108229575A (en) * | 2018-01-19 | 2018-06-29 | 百度在线网络技术（北京）有限公司 | For detecting the method and apparatus of target |
Non-Patent Citations (2)
| Title |
|---|
| 侯纯: "不良图像检测系统的设计与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
| 陈骁: "基于感兴趣区域检测的网络不良图片识别研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110310142A (en) * | 2019-05-20 | 2019-10-08 | 浙江新再灵科技股份有限公司 | Elevator card put-on method and device based on crowd's value analysis |
| CN111814852A (en) * | 2020-06-24 | 2020-10-23 | 理光软件研究所（北京）有限公司 | Image detection method, image detection device, electronic equipment and computer-readable storage medium |
| CN113454649A (en) * | 2021-06-17 | 2021-09-28 | 商汤国际私人有限公司 | Target detection method, target detection device, electronic equipment and computer-readable storage medium |
| WO2022263904A1 (en) * | 2021-06-17 | 2022-12-22 | Sensetime International Pte. Ltd. | Target detection methods, apparatuses, electronic devices and computer-readable storage media |
| AU2021204563A1 (en) * | 2021-06-17 | 2023-01-19 | Sensetime International Pte. Ltd. | Target detection methods, apparatuses, electronic devices and computer-readable storage media |
| CN113454649B (en) * | 2021-06-17 | 2024-05-24 | 商汤国际私人有限公司 | Target detection method, apparatus, electronic device, and computer-readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2020006964A1 (en) | 2020-01-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109558779A (en) | Image detecting method and device | |
| CN108985208A (en) | The method and apparatus for generating image detection model | |
| US11487995B2 (en) | Method and apparatus for determining image quality | |
| CN108427939B (en) | Model generation method and device | |
| CN108038880B (en) | Method and apparatus for processing image | |
| CN108510472B (en) | Method and apparatus for handling image | |
| CN108154196B (en) | Method and apparatus for exporting image | |
| CN109815156A (en) | Displaying test method, device, equipment and the storage medium of visual element in the page | |
| CN108898086A (en) | Method of video image processing and device, computer-readable medium and electronic equipment | |
| CN109389068A (en) | The method and apparatus of driving behavior for identification | |
| CN109002842A (en) | Image-recognizing method and device | |
| CN109410218A (en) | Method and apparatus for generating vehicle damage information | |
| CN108830235A (en) | Method and apparatus for generating information | |
| CN109446990A (en) | Method and apparatus for generating information | |
| CN108494778A (en) | Identity identifying method and device | |
| CN109034069A (en) | Method and apparatus for generating information | |
| CN109359676A (en) | Method and apparatus for generating vehicle damage information | |
| CN109308490A (en) | Method and apparatus for generating information | |
| CN109086780A (en) | Method and apparatus for detecting electrode piece burr | |
| CN110443824A (en) | Method and apparatus for generating information | |
| CN109977839A (en) | Information processing method and device | |
| CN108446659A (en) | Method and apparatus for detecting facial image | |
| CN109871791A (en) | Image processing method and device | |
| CN108509921A (en) | Method and apparatus for generating information | |
| CN108449514A (en) | Information processing method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190402 |
|
| RJ01 | Rejection of invention patent application after publication |