CN116958795A - Method and device for identifying flip image, electronic equipment and storage medium - Google Patents
Method and device for identifying flip image, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN116958795A CN116958795A CN202310799255.8A CN202310799255A CN116958795A CN 116958795 A CN116958795 A CN 116958795A CN 202310799255 A CN202310799255 A CN 202310799255A CN 116958795 A CN116958795 A CN 116958795A
- Authority
- CN
- China
- Prior art keywords
- picture
- target picture
- flip
- target
- probability
- 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
- 238000000034 method Methods 0.000 title claims abstract description 101
- 238000013145 classification model Methods 0.000 claims abstract description 40
- 230000008569 process Effects 0.000 claims abstract description 34
- 238000012545 processing Methods 0.000 claims description 44
- 238000000605 extraction Methods 0.000 claims description 17
- 230000011218 segmentation Effects 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 238000003384 imaging method Methods 0.000 claims description 6
- 238000003909 pattern recognition Methods 0.000 claims description 3
- 238000012549 training Methods 0.000 description 22
- 230000009471 action Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 4
- 230000008878 coupling Effects 0.000 description 3
- 238000010168 coupling process Methods 0.000 description 3
- 238000005859 coupling reaction Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000004091 panning Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/95—Pattern authentication; Markers therefor; Forgery detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/54—Extraction of image or video features relating to texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/35—Categorising the entire scene, e.g. birthday party or wedding scene
- G06V20/36—Indoor scenes
Abstract
The application provides a method and a device for identifying a flip image, electronic equipment and a storage medium, and relates to the field of image identification. The identification method of the flip image comprises the following steps: acquiring a target picture based on a house indoor scene classification model, wherein the target picture is a house indoor scene picture; executing a reproduction identification process on the target picture, wherein the reproduction identification process comprises the following steps: and determining that the target picture is a flip picture when the flip probability of the target picture is larger than a probability threshold value or the picture identification result of the target picture indicates that a flip mark exists. The method and the device improve the accuracy of the reproduction identification of the house picture by utilizing a plurality of characteristic identifications aiming at the identified house indoor scene picture.
Description
Technical Field
The present application relates to the field of image recognition, and in particular, to a method and apparatus for recognizing a flip image, an electronic device, and a storage medium.
Background
With the popularization of the internet, people start to distribute images such as pictures and videos on the network for entertainment, work and the like. Thus, a phenomenon of flipping an image issued by another person occurs. With the increase of such phenomena, a technique of recognizing a reproduction image has also been developed.
In the current method for identifying the flip, the single feature of the flip image is usually used for identifying the flip behavior. For example, by identifying whether the anti-theft watermark is present on the image, the image is determined to be a flip image in the case that the anti-theft watermark is identified from the image; and in the case that the anti-theft watermark is not recognized from the image, judging that the image is a non-flipped image.
However, for scenes such as house source authentication with house pictures, house source design with house pictures, etc., the image involved may have multiple scenes, a large change in shooting angle, even a case where the real features of the shooting object are fused with the features of the flipped image, etc. Aiming at the complex image of the shooting object, the mode of performing the shooting identification by using the single characteristic of the image is adopted to perform the shooting identification, and the accuracy of the shooting identification is lower possibly because the anti-theft condition reflected by the single identification characteristic is inaccurate.
Disclosure of Invention
In view of the above, the present application provides a method, an apparatus, an electronic device, and a storage medium for identifying a flipped image, which improve the accuracy of flipped identification to a certain extent by using the features of a plurality of flipped images.
According to a first aspect of the present application, there is provided a method of identifying a flip image, the method comprising:
acquiring a target picture based on a house indoor scene classification model, wherein the target picture is a house indoor scene picture;
executing a reproduction identification process on the target picture, wherein the reproduction identification process comprises the following steps: and determining that the target picture is a flip picture when the flip probability of the target picture is larger than a probability threshold value or the picture identification result of the target picture indicates that a flip mark exists.
According to a second aspect of the present application, there is provided an apparatus for identifying a flip image, the apparatus comprising:
the acquisition module is used for acquiring a target picture based on the house indoor scene classification model, wherein the target picture is a house indoor scene picture;
the device comprises a reproduction identification processing module, a reproduction identification processing module and a reproduction identification processing module, wherein the reproduction identification processing module is used for executing reproduction identification processing on the target picture, and the reproduction identification processing comprises: and determining that the target picture is a flip picture when the flip probability of the target picture is larger than a probability threshold value or the picture identification result of the target picture indicates that a flip mark exists.
According to a third aspect of the present application, there is provided an electronic device comprising: a processor, a memory and a computer program stored on the memory and executable on the processor, which when executed by the processor performs the steps of the method for identifying a flipped image as described in any of the first aspects.
According to a fourth aspect of the present application there is provided a computer readable storage medium having stored thereon a computer program which when executed by a processor implements the steps of the method of identifying a flip image as described in any of the first aspects.
According to a fifth aspect of the present application, there is provided an electronic device comprising: a processor, a memory and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method for identifying a flipped image as described in any of the first aspects when the program is executed.
Aiming at the related technology, the application has the following advantages:
according to the method, the device and the electronic equipment for identifying the flipped pictures, the flipped identification processing is carried out on the house indoor scene pictures acquired based on the house indoor scene classification model, so that when the flipped probability of the house indoor scene pictures is larger than the probability threshold, or the picture identification result of the house indoor scene pictures indicates that the flipped mark exists, the target pictures are determined to be flipped pictures. According to the technical scheme, aiming at the identified indoor scene pictures of the house, the two features of the picture, namely the probability of the picture to be flipped and the existence condition of the flipped mark in the picture, are respectively utilized to jointly carry out flipped identification of the picture.
Drawings
FIG. 1 is a flowchart of a method for identifying a flip image according to an embodiment of the present application;
FIG. 2 is a flowchart of another method for identifying a flipped image according to an embodiment of the present application;
FIG. 3 is a flowchart of a method for identifying a flipped image according to another embodiment of the present application;
FIG. 4 is a flowchart of another method for identifying a flipped image according to another embodiment of the present application;
FIG. 5 is a block diagram of an apparatus for recognizing a flip image according to an embodiment of the present application;
fig. 6 is a block diagram of an electronic device according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
Referring to fig. 1, a flowchart of a method for identifying a flipped image according to an embodiment of the present application is shown. The identification method can be applied to electronic equipment. Optionally, the electronic device may be a computer, a mobile phone, a tablet, a server, a cloud server, a server cluster, or the like. As shown in fig. 1, the method for identifying a flip image may include:
And step 101, acquiring a target picture based on the house indoor scene classification model. The target picture is a house indoor scene picture.
Optionally, the target picture may be a target picture obtained by screening the house indoor scene classification model from the pictures to be identified. The house indoor scene classification model is used for identifying whether the input picture is a house indoor scene picture. The picture to be identified may be a picture to be identified uploaded by the user, or the picture to be identified may also be a house picture saved by the user on the target interface. Alternatively, the electronic device may obtain a locally stored picture to be identified. Of course, the electronic device may also acquire a picture to be identified from a third party device, and so on.
In an alternative implementation, the process of acquiring the target picture based on the house indoor scene classification model may include: and inputting the picture to be identified into a house indoor scene classification model to obtain a classification result corresponding to the picture to be identified. The classification result indicates whether the picture to be identified is a house indoor scene picture.
And if the classification result indicates that the picture to be identified is an indoor scene picture, taking the picture to be identified as a target picture. If the classification result indicates that the picture to be identified is not the indoor scene picture, the picture to be identified is not taken as the target picture.
Alternatively, the house indoor scene classification model may be obtained by training a pre-constructed network by using a third training set. The third training set comprises a plurality of training pictures and labels corresponding to the training pictures. The tag is used for indicating whether the training picture is a house indoor scene picture. The plurality of training pictures includes house indoor scene pictures and non-indoor scene pictures. For example, the non-indoor scene pictures may be house cell wide field pictures, house outside floor pictures, and the like. Illustratively, the house indoor scene classification model may be ResNeXt, inceptionV or MobileNet or the like.
And 102, executing the flip identification processing on the target picture. The reproduction identification process comprises the following steps: and determining that the target picture is a flipped picture when the flipped probability of the target picture is larger than a probability threshold value or the picture identification result of the target picture indicates that a flipped mark exists.
In some embodiments of the present application, the electronic device may determine that the target picture is a non-flipped picture when the probability of flipping the target picture is less than or equal to the probability threshold, and the recognition result of the target picture indicates that the target picture does not have a flipped mark. The probability of the image is the probability that the target image is the image. The picture identification result indicates whether a flip flag exists in the target picture.
Optionally, the probability of panning is output data obtained by inputting the target picture into the moire recognition model. The probability of a snap shot indicates a probability of determining that the target picture is a snap shot picture based on moire features of the target picture.
In some embodiments of the present application, a moire recognition model may be used to extract moire features of a target picture, and determine whether the target picture is a flip picture based on the moire features. In an alternative implementation, the moire recognition model may be obtained by training a pre-built network model using a first training set. The first training set includes: the system comprises a plurality of training pictures and labels corresponding to the training pictures. The tag is used for indicating whether the training picture is a flip picture or not. The plurality of training pictures comprise shooting original pictures and flip pictures with mole patterns. Shooting an original image refers to shooting an obtained picture of a real scene by adopting shooting equipment.
In an alternative implementation, the moire recognition model may include a moire extraction model and a classification model. The moire extraction model is used for outputting moire features extracted from an input picture with respect to the picture. The classification model is used to output a mole pattern confidence score based on the entered mole pattern features. The moire confidence score, i.e., the probability of a tap, is used to indicate the probability of having a moire in the input picture. Since the flip pictures typically have moire. Thus, the greater the probability that a picture has moire, the greater the probability that the picture is a flipped picture. Based on this, the moire confidence score is also used to indicate whether the input picture is a flipped picture. Illustratively, the moire recognition model may be a ShuffleNet or an EfficientNet, or the like.
Further alternatively, the picture recognition result is output data obtained by inputting the target picture into the picture recognition model. The picture identification result indicates whether a flip flag exists in the target picture.
The picture identification model is used for identifying whether an input picture has a reproduction mark or not. The picture-turning mark is used for indicating that the picture has a high probability of picture-turning. Alternatively, the tap mark may be a watermark mark, an object mark of an object that may be involved in the tap action, or the like. For example, the tap mark is a mouse arrow image, a lens image, or the like. In an alternative implementation, the image recognition model may be obtained by training a pre-built network model using a second training set. The second training set comprises a plurality of training pictures and labels corresponding to the training pictures. The tag is used for indicating whether the training picture has a flap mark or not. The plurality of training pictures includes pictures without a roll-over flag and pictures with a roll-over flag. The reproduction marks on the training pictures can be various different reproduction marks. The picture recognition model may be a neural network model such as a DenseNet model or a ResNet model, for example.
For example, after the electronic device obtains the target picture, the target picture may be input to the moire recognition model to obtain the probability of flipping. And under the condition that the probability of the flip is larger than the probability threshold, determining that the target picture is the flip picture. And under the condition that the probability of the flip is smaller than or equal to the probability threshold, inputting the target picture into the picture recognition model to obtain a picture recognition result. And under the condition that the picture identification result indicates that a flip mark exists in the target picture, determining that the target picture is a flip picture. And under the condition that the picture identification result indicates that the reproduction mark does not exist in the target picture, determining that the target picture is a non-reproduction picture.
Of course, the electronic device may input the target picture into the picture recognition model to obtain the picture recognition result after the target picture is obtained. And under the condition that the image identification result indicates that the target image does not have the reproduction mark, inputting the target image into the mole pattern identification model to obtain the reproduction probability. And determining whether the target picture is a flipped picture or not based on the flipped probability and the probability threshold. Or the electronic equipment can also input the target picture into the picture recognition model and the mole pattern recognition model simultaneously to obtain a picture recognition result and a flip probability. The electronic device determines whether the target picture is a flip picture based on the picture identification result, the flip probability and the probability threshold.
In the embodiment of the application, the image of the house indoor scene acquired based on the house indoor scene classification model is subjected to the image-turning identification process, so that the target image is determined to be the image-turning image when the image-turning probability of the image of the house indoor scene is larger than the probability threshold value or the image identification result of the image of the house indoor scene indicates that the image-turning mark exists. According to the technical scheme, aiming at the identified indoor scene pictures of the house, the two features of the picture, namely the probability of the picture to be flipped and the existence condition of the flipped mark in the picture, are respectively utilized to jointly carry out flipped identification of the picture.
Further, when the probability of the turning over indicates that the target picture is the probability of the turning over picture based on the mole pattern feature of the target picture and the picture identification result indicates whether the turning over mark exists in the target picture, the turning over identification of the picture is carried out by respectively utilizing the mole pattern feature and the turning over mark feature possibly existing in the turning over picture through adopting the mole pattern identification model and the picture identification model.
In an alternative implementation, the processing of identifying the flip performed by the electronic device on the target picture may further include: and determining that the target picture is the flipped picture when the flipped probability of the target picture is larger than a probability threshold, or the picture identification result of the target picture indicates that a flipped mark exists, or the area of the equipment image area in the target picture is larger than an area threshold.
In some embodiments of the present application, the electronic device may determine that the target picture is a non-flipped picture when the probability of flipping the target picture is less than or equal to the probability threshold, and the recognition result of the target picture indicates that the target picture does not have a flipped mark, and the area of the device image area in the target picture is less than or equal to the area threshold.
The device image area is an imaging area of the flip device in the target picture. When a user uses the photographing device to photograph a picture, at least part of the photographing device may fall into an imaging area of the photographing device or be reflected by the photographed picture, so that at least part of the photographing device images on a target picture. The electronic device may identify whether the target picture has a device image area in order to read an area of the device image area in case that the device image area is present in the target picture. In one possible scenario, when a user uses a flipping device to flip a picture, the device frame of the flipping device more easily falls into the imaging area of the flipping device, or is reflected by the flipped picture, so that the device frame of the flipping device images on the target picture. Thus, the device image area of the target picture may be an imaging area of the device bezel of the flip device in the target picture.
Alternatively, the area of the device image area may be the pixel area of the device image area on the target picture. The electronic device may input the target picture into the detection segmentation model to obtain a device picture (i.e., a mask of the device) of the device image area in the target picture. The pixel area of the device picture, i.e. the area of the device image area, is calculated. And under the condition that the pixel area is larger than the area threshold value, determining that the target picture is a flip picture.
The detection segmentation model is used for identifying the device image area in the input picture and segmenting and outputting the picture of the device image area in the input picture. Alternatively, the area threshold may be proportional to the picture area of the target picture. For example, the area threshold may be a product of a preset ratio and a picture area of the target picture. The preset ratio is the ratio of the area of the preset equipment image area to the picture area of the target picture. In an alternative implementation, the pixel area of a picture may be the number of pixels on the picture. The implementation process of the pixel area of the electronic device computing device picture can comprise the following steps: the electronic device can count the number of pixel points in the device picture to obtain a pixel area, and the pixel area is used as an area. Correspondingly, the area threshold is the product of the preset ratio and the number of pixel points on the target picture. The preset ratio may be a ratio of a preset device image area to the number of pixels of the target picture.
For example, as shown in fig. 2, after the electronic device acquires the target picture, the electronic device may input the target picture into the moire recognition model to obtain the probability of flipping. And judging whether the probability of the flick is larger than a probability threshold. And under the condition that the probability of the flip is larger than the probability threshold, determining that the target picture is the flip picture. And under the condition that the probability of the flip is smaller than or equal to the probability threshold value, inputting the target picture into the detection segmentation model to obtain the equipment picture of the equipment image area in the target picture. And the electronic equipment calculates the pixel area of the equipment picture to obtain the area of the area. And judging whether the area of the region is larger than an area threshold value. Under the condition that the area of the region is larger than the area threshold value, determining that the target picture is a flip picture; and under the condition that the area of the region is smaller than or equal to the area threshold value, inputting the target picture into a picture recognition model to obtain a picture recognition result. And the electronic equipment determines that the target picture is the flip picture under the condition that the picture identification result indicates that the flip mark exists. And the electronic equipment determines that the target picture is a non-flipped picture under the condition that the picture identification result indicates that the flipped mark does not exist.
The probability of existence of the moire feature, the probability of existence of the equipment image area and the probability of existence of the turner mark in the turner picture are sequentially reduced. Therefore, in the process of identifying the target picture by turning over, the first mole pattern classification model, the picture identification model and the last identification sequence of the detection segmentation model are adopted preferentially, so that the characteristic with larger probability existing in the picture by turning over can be utilized preferentially in the process of identifying the picture by turning over, the identification speed of the picture by turning over is improved, the processing steps of single-time picture by turning over identification are reduced as much as possible, and the identification efficiency of the picture by turning over is improved.
It should be noted that, after the electronic device obtains the target picture to be identified, the target picture may be input into the picture identification model to obtain the picture identification result. And under the condition that the image identification result indicates that the target image does not have the reproduction mark, inputting the target image into the mole pattern identification model to obtain the reproduction probability. And under the condition that the probability of the flip is smaller than or equal to the probability threshold value, inputting the target picture into the detection segmentation model to obtain the equipment picture of the equipment image area in the target picture. And the electronic equipment calculates the pixel area of the equipment picture to obtain the area of the area. And under the condition that the area of the region is larger than the area threshold value, determining that the target picture is a flip picture. Of course, the electronic device may also adopt the identification sequence of detecting the segmentation model, then adopting the mole pattern classification model, and finally adopting the picture identification model for the flip identification processing of the target picture. Or firstly adopting a detection segmentation model, then adopting a picture recognition model, and finally adopting a recognition sequence of a mole pattern classification model. The recognition sequence of the three models is not particularly limited in the embodiment of the application.
In summary, according to the method for identifying a flipped image provided by the embodiment of the application, flipped identification processing is performed on a house indoor scene picture acquired based on a house indoor scene classification model, so that when the flipped probability of the house indoor scene picture is greater than a probability threshold, or if a picture identification result of the house indoor scene picture indicates that a flipped mark exists, it is determined that the target picture is a flipped picture. According to the technical scheme, the two characteristics of the image reproduction probability and the existence condition of the reproduction mark in the image are respectively utilized to jointly carry out reproduction recognition of the image, and compared with the method for carrying out reproduction recognition by single characteristic in the related art, the method for recognizing the house indoor scene image by utilizing the multiple characteristic reproduction recognition scheme improves the accuracy of reproduction recognition of the house image.
Further, when the probability of the turning over indicates that the target picture is the probability of the turning over picture based on the mole pattern feature of the target picture and the picture identification result indicates whether the turning over mark exists in the target picture, the turning over identification of the picture is carried out by respectively utilizing the mole pattern feature and the turning over mark feature possibly existing in the turning over picture through adopting the mole pattern identification model and the picture identification model.
Referring to fig. 3, a flowchart of another method for identifying a flipped image according to an embodiment of the present application is shown. The identification method can be applied to electronic equipment. Optionally, the electronic device may be a computer, a mobile phone, a tablet, a server, a cloud server, a server cluster, or the like. As shown in fig. 3, the method for identifying the flipped image may include:
step 301, taking a plurality of video frames in the video to be identified as a plurality of pictures to be identified.
Alternatively, the plurality of video frames in the video may refer to all of the video frames in the video. Or, part of the video frames in the video. Under the condition that the plurality of video frames are part of video frames in the video, the electronic equipment can perform frame extraction processing on the video to be identified to obtain the plurality of video frames. In an alternative implementation manner, the process that the electronic device performs frame extraction processing on the video to be identified to obtain a plurality of video frames may include: the electronic equipment takes a plurality of video frames randomly extracted from the video to be identified as a plurality of pictures to be identified. Or, the process of performing frame extraction processing on the video to be identified by the electronic device to obtain a plurality of video frames may include: the electronic equipment can determine the needle extraction rate based on the motion characteristics of the video to be identified, and perform frame extraction processing on the video according to the needle extraction rate to obtain a plurality of video frames. The pin extraction rate is the frame extraction frequency.
And 302, acquiring a plurality of target pictures from a plurality of pictures to be identified based on the house indoor scene classification model.
Optionally, the electronic device may screen house indoor scene pictures in the plurality of pictures to be identified using house indoor scene classification models. The electronic device sequentially inputs the pictures to be identified into the house indoor scene classification model to obtain classification results corresponding to the pictures to be identified. And indicating the pictures to be identified of the indoor scene pictures as target pictures according to the classification result in the multiple pictures to be identified, and obtaining multiple target pictures. It should be noted that there is a case where one target picture is obtained from a plurality of pictures to be identified based on a house indoor scene classification model. In this case, the electronic device may perform a flip identification process on the target picture to determine whether the target picture is a flip picture.
Step 303, performing a roll-over identification process on each of the plurality of target pictures.
The specific process of the process of identifying the beats may refer to the description of the process of identifying the beats in the foregoing embodiment, which is not described in detail in the embodiment of the present application. For example, the process of identifying a flip includes at least the step 102 of determining that the target picture is a flip picture when the probability of the flip of the target picture is greater than the probability threshold, or when the picture identification result of the target picture indicates that a flip flag exists. Of course, in an alternative implementation, the process of the tap recognition process further includes: and determining that the target picture is the flipped picture when the flipped probability of the target picture is larger than a probability threshold, or the picture identification result of the target picture indicates that a flipped mark exists, or the area of the equipment image area in the target picture is larger than an area threshold.
And step 304, determining that the video is a flip video under the condition that the ratio of the first picture to the plurality of target pictures is greater than a ratio threshold. The first picture is a target picture determined to be a flip picture.
Optionally, after performing the flip recognition processing on all the target pictures, the electronic device may count the number of the first pictures in all the target pictures, and further calculate a ratio of the number of the first pictures to the total number of the target pictures. It is determined whether the ratio is greater than a ratio threshold. And under the condition that the ratio is smaller than or equal to the ratio threshold value, the probability that the video is the original video is larger, and the video is determined to be the non-flip video. And under the condition that the ratio is larger than the ratio threshold, indicating that more turnup pictures are in the video, and determining that the video is the turnup video. Illustratively, the ratio threshold may be 50%, 70%, 80%, or the like.
In the embodiment of the application, the overturn identification processing is respectively carried out on a plurality of house indoor scene pictures in the video to be identified to obtain the overturn identification result of each house indoor scene picture, so as to determine the overturn identification result of the video based on the duty ratio of the overturn pictures in the plurality of house indoor scene pictures, and realize the overturn identification of the video. Particularly, aiming at the video mixed by the real scene and the flip scene, the flip video frames in the video, namely the flip pictures, can be selected with high probability by performing frame extraction processing on the video to be identified, so that the accuracy of the video flip identification is ensured to a certain extent. It will be appreciated that the higher the frame extraction rate in the frame extraction process, the higher the accuracy of the video rollover recognition.
And, because compared with the outdoor scene picture of the house, the difficulty of adopting the non-flapping means to acquire the indoor scene picture of the house is larger. Therefore, compared with the outdoor scene pictures, the indoor scene pictures of the house have higher probability of being flipped, and further, the efficiency of video flipping recognition is higher based on the flipping recognition result of the video frames describing the indoor scenes in the video. In particular, for a house original video use scene such as house authenticity authentication by using house videos, the house indoor scene picture contributes more to house authenticity judgment than the house outdoor scene picture. Therefore, the accuracy of judging the authenticity of the house based on the video flip recognition result of the house indoor scene picture is high.
The following examples are taken as examples of the embodiment of the application, and the method for identifying the flipped image provided by the application is further schematically illustrated. As shown in fig. 4, the electronic device performs frame extraction processing on the video to be identified, so as to obtain n video frames. The video frame is in a picture format, and n is a positive integer. The electronic device sets initial values of the identified frame number, the flipped video frame number, the non-flipped video frame number and the indoor video frame number to 0. Wherein the identified frame number records the number of video frames for which the tap identification process has been performed. The roll-over video frame number record is determined as the number of video frames of the roll-over picture through the roll-over identification process. The non-flipped video frame number record determines the number of video frames that are non-flipped pictures through a flipped identification process. The indoor video frame number records the classification result output by the house indoor scene classification model is determined as the number of video frames of the indoor scene picture (i.e., target picture).
The electronic device increments the value of the number of frames identified by 1. The electronic equipment inputs any one first video frame in the plurality of video frames into a house indoor scene classification model to obtain a classification result corresponding to the first video frame. The first video frame is a video frame of the plurality of video frames that is not input to the house indoor scene classification model.
The electronic device judges whether the classification result indicates that the first video frame is an indoor scene picture. And if the classification result indicates that the first video frame is not an indoor scene picture, the electronic device judges whether the number of the processed recognized frames is less than n.
And when the processed identified frame number is smaller than n, the electronic equipment returns to execute the addition of 1 to the value of the identified frame number, and any one first video frame in the plurality of video frames is input into the house indoor scene classification model to obtain a classification result corresponding to the first video frame.
And under the condition that the processed identified frame number is greater than or equal to n, the electronic equipment calculates the ratio of the frame number of the turnup video to the frame number of the indoor video to obtain the proportion of the turnup video, namely the ratio of the first picture to the plurality of target pictures. The electronic device judges whether the proportion of the flipped video is larger than a proportion threshold value. Under the condition that the proportion of the video to be shot is larger than a proportion threshold value, determining that the video is the video to be shot; and under the condition that the proportion of the flipped video is smaller than or equal to the proportion threshold value, the electronic equipment determines that the video is a non-flipped video.
And under the condition that the classification result indicates that the first video frame is an indoor scene picture, the first video frame is a target picture. And the electronic equipment performs 1 adding processing on the indoor video frame number, and inputs the first video frame into a mole pattern extraction model to obtain mole pattern characteristics. And the electronic equipment inputs the moire characteristics into the classification model to obtain a moire confidence score, namely the probability of the flipping. And judging whether the moire confidence score is larger than a moire threshold, namely a flap threshold.
And under the condition that the mole pattern confidence score is larger than the mole pattern threshold, the electronic equipment determines that the first video frame is a flip picture, adds 1 to the value of the flip video frame number, and returns to whether the identified frame number after the judgment processing is smaller than n. And under the condition that the moire confidence score is smaller than or equal to the moire threshold value, the electronic equipment inputs the first video frame into the detection segmentation model to obtain an equipment picture of the equipment image area in the first picture. The electronic device determines whether the pixel area of the device picture is greater than an area threshold.
Under the condition that the pixel area of the equipment picture is larger than the area threshold value, the electronic equipment determines that the first video frame is a flip picture, the value of the flip video frame number is added with 1, and the electronic equipment returns to judge whether the identified frame number after the processing is smaller than n. And under the condition that the pixel area of the equipment picture is smaller than or equal to the area threshold value, the electronic equipment inputs the first video frame into the picture identification model to obtain a picture identification result.
The electronic equipment judges whether the picture identification result indicates that a flipping mark exists. And under the condition that the picture identification result indicates that a reproduction mark exists, the electronic equipment determines that the first video frame is a reproduction picture, adds 1 to the value of the reproduction video frame number, and returns to whether the identified frame number after the judgment processing is smaller than n.
And under the condition that the picture identification result indicates that the reproduction mark does not exist, the electronic equipment determines that the first video frame is a non-reproduction picture, and the value of the non-reproduction video frame number is increased by 1. The electronic device returns to whether the number of recognized frames after the execution of the judgment processing is smaller than n.
It should be noted that the content of the method embodiments of the present application may be referred to each other. For example, the relevant interpretation of the classification model for the house indoor scene in the method embodiment shown in fig. 4 may refer to the relevant interpretation of the classification model for the house indoor scene in the method embodiment shown in fig. 3 described above.
In summary, according to the method for identifying the flipped image provided by the embodiment of the application, the flipped image identification process is performed on the house indoor scene image acquired based on the house indoor scene classification model, so that the target image is determined to be the flipped image when the flipped image probability of the house indoor scene image is greater than the probability threshold, or the image identification result of the house indoor scene image indicates that the flipped image is present. According to the technical scheme, the two characteristics of the picture flip probability and the existence condition of the flip mark in the picture are respectively utilized to jointly perform the picture flip recognition, and compared with the method for recognizing the house indoor scene picture by single characteristic flip recognition in the related art, the method for recognizing whether the house indoor scene picture is the flip picture or not by utilizing the plurality of characteristics improves the accuracy of the flip video/picture recognition.
Fig. 5 is a block diagram of an apparatus for recognizing a flipped image according to an embodiment of the present application. As shown in fig. 5, the apparatus 500 for recognizing a flip image includes: an acquisition module 501 and a flap recognition processing module 502.
The obtaining module 501 is configured to obtain a target picture based on a house indoor scene classification model, where the target picture is a house indoor scene picture;
the reproduction identification processing module 502 is configured to perform reproduction identification processing on the target picture, where the reproduction identification processing includes: and determining that the target picture is a flipped picture when the flipped probability of the target picture is larger than a probability threshold value or the picture identification result of the target picture indicates that a flipped mark exists.
Optionally, the probability of flipping is output data obtained by inputting the target picture into a moire recognition model, the probability of flipping indicates a probability of determining that the target picture is a flipped picture based on the moire feature of the target picture, the picture recognition result is output data obtained by inputting the target picture into a picture recognition model, and the picture recognition result indicates whether a flipped mark exists in the target picture.
Optionally, the flap recognition processing module 502 is further configured to: and determining that the target picture is a flip picture when the flip probability of the target picture is greater than a probability threshold, or the picture identification result of the target picture indicates that a flip mark exists, or the area of an equipment image area in the target picture is greater than an area threshold, wherein the equipment image area is an imaging area of the flip equipment in the target picture.
Optionally, the flap recognition processing module 502 is further configured to:
inputting the target picture into a mole pattern recognition model to obtain a reproduction probability;
under the condition that the probability of the turning over is larger than a probability threshold, determining that the target picture is the turning over picture, and under the condition that the probability of the turning over is smaller than or equal to the probability threshold, inputting the target picture into a detection segmentation model to obtain a device picture of a device image area in the target picture;
calculating the pixel area of the equipment picture to obtain the area of the area;
under the condition that the area of the area is larger than the area threshold value, determining that the target picture is a flip picture, and under the condition that the area of the area is smaller than or equal to the area threshold value, inputting the target picture into a picture recognition model to obtain a picture recognition result;
and under the condition that the picture identification result indicates that a flip mark exists, determining that the target picture is a flip picture.
Optionally, the obtaining module 501 is further configured to:
inputting the picture to be identified into a house indoor scene classification model to obtain a classification result corresponding to the picture to be identified, wherein the classification result indicates whether the picture to be identified is a house indoor scene picture or not; and if the classification result indicates that the picture to be identified is an indoor scene picture, taking the picture to be identified as a target picture.
Optionally, the obtaining module 501 is further configured to: taking a plurality of video frames in the video to be identified as a plurality of pictures to be identified; acquiring a plurality of target pictures from a plurality of pictures to be identified based on a house indoor scene classification model;
the reproduction identification processing module 502 is further configured to perform reproduction identification processing on each of the plurality of target pictures; and the method is also used for determining that the video is a flip video under the condition that the ratio of the first picture to the plurality of target pictures is larger than a proportion threshold value, and the first picture is the target picture determined as the flip picture. .
Optionally, the obtaining module 501 is further configured to perform frame extraction processing on the video to be identified to obtain a plurality of video frames.
In summary, according to the apparatus for identifying a flipped image provided by the embodiment of the present application, by performing flipped identification processing on a house indoor scene image obtained based on a house indoor scene classification model, when the probability of flipped of the house indoor scene image is greater than a probability threshold, or when a picture identification result of the house indoor scene image indicates that a flipped mark exists, it is determined that the target image is a flipped image. According to the technical scheme, the two characteristics of the picture flip probability and the existence condition of the flip mark in the picture are respectively utilized to jointly perform the picture flip recognition, and compared with the method for recognizing the house indoor scene picture by single characteristic flip recognition in the related art, the method for recognizing whether the house indoor scene picture is the flip picture or not by utilizing the plurality of characteristics improves the accuracy of the flip video/picture recognition.
The device for identifying the flipped image provided by the embodiment of the application is provided with the corresponding functional module for executing the method for identifying the flipped image, can execute any of the methods for identifying the flipped image provided by the embodiment of the application, and can achieve the same beneficial effects.
In still another embodiment of the present application, there is also provided an electronic device, which may include: the image reproduction device comprises a processor, a memory and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes each process of the embodiment of the image reproduction identification method when executing the program, and can achieve the same technical effect, and the repetition is avoided, so that the description is omitted.
As illustrated in fig. 6, the electronic device may specifically include: processor 601, storage device 602, display 603, input device 604, output device 605, and communication device 606. The number of processors 601 in the electronic device may be one or more, one processor 601 being taken as an example in fig. 6. The processor 601, storage 602, display 603, input 604, output 605, and communication 606 of the electronic device may be connected by a bus or other means.
In yet another embodiment of the present application, a computer readable storage medium is provided, where instructions are stored, which when executed on a computer, cause the computer to perform the method for identifying a flipped image according to any of the above embodiments.
In yet another embodiment of the present application, a computer program product containing instructions that, when run on a computer, cause the computer to perform the method for identifying a flipped image as described in any of the above embodiments is also provided.
It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
From the above description of the embodiments, it will be clear to those skilled in the art that the above-described embodiment method may be implemented by means of software plus a necessary general hardware platform, but of course may also be implemented by means of hardware, but in many cases the former is a preferred embodiment. Based on such understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art in the form of a software product stored in a storage medium (e.g. ROM/RAM, magnetic disk, optical disk) comprising instructions for causing a terminal (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to perform the method according to the embodiments of the present invention.
The embodiments of the present invention have been described above with reference to the accompanying drawings, but the present invention is not limited to the above-described embodiments, which are merely illustrative and not restrictive, and many forms may be made by those having ordinary skill in the art without departing from the spirit of the present invention and the scope of the claims, which are to be protected by the present invention.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, and are not repeated herein.
In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of the units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or units, which may be in electrical, mechanical or other form.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present application may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit.
The functions, if implemented in the form of software functional units and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, a server, a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a usb disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk, etc.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (10)
1. A method for identifying a flip image, the method comprising:
acquiring a target picture based on a house indoor scene classification model, wherein the target picture is a house indoor scene picture;
executing a reproduction identification process on the target picture, wherein the reproduction identification process comprises the following steps: and determining that the target picture is a flip picture when the flip probability of the target picture is larger than a probability threshold value or the picture identification result of the target picture indicates that a flip mark exists.
2. The method according to claim 1, wherein the flip probability is output data obtained by inputting the target picture into a moire recognition model, the flip probability indicates a probability of determining that the target picture is a flip picture based on moire characteristics of the target picture, the picture recognition result is output data obtained by inputting the target picture into a picture recognition model, and the picture recognition result indicates whether or not the flip flag is present in the target picture.
3. The method according to claim 1, wherein the determining that the target picture is a flipped picture if the flipped probability of the target picture is greater than a probability threshold, or if a picture identification result of the target picture indicates that a flipped flag is present, comprises:
And determining that the target picture is a flip picture when the flip probability of the target picture is greater than a probability threshold, or the picture identification result of the target picture indicates that a flip mark exists, or the area of an equipment image area in the target picture is greater than an area threshold, wherein the equipment image area is an imaging area of a flip equipment in the target picture.
4. A method according to claim 3, wherein the determining that the target picture is a flipped picture if the probability of flipping the target picture is greater than a probability threshold, or if a picture identification result of the target picture indicates that a flipped flag is present, or if an area of an image area of a device in the target picture is greater than an area threshold, comprises:
inputting the target picture into a mole pattern recognition model to obtain the reproduction probability;
under the condition that the probability of the turning over is larger than a probability threshold, determining that the target picture is the turning over picture, and under the condition that the probability of the turning over is smaller than or equal to the probability threshold, inputting the target picture into a detection segmentation model to obtain a device picture of a device image area in the target picture;
Calculating the pixel area of the equipment picture to obtain the area of the area;
determining that the target picture is a flip picture when the area of the region is larger than an area threshold value, and inputting the target picture into a picture recognition model when the area of the region is smaller than or equal to the area threshold value to obtain the picture recognition result;
and under the condition that the picture identification result indicates that a turner mark exists, determining that the target picture is a turner picture.
5. The method according to any one of claims 1 to 4, wherein the acquiring the target picture based on the house indoor scene classification model includes:
inputting a picture to be identified into a house indoor scene classification model to obtain a classification result corresponding to the picture to be identified, wherein the classification result indicates whether the picture to be identified is a house indoor scene picture or not;
and if the classification result indicates that the picture to be identified is an indoor scene picture, taking the picture to be identified as the target picture.
6. The method according to any one of claims 1 to 4, wherein the acquiring the target picture based on the house indoor scene classification model includes:
Taking a plurality of video frames in the video to be identified as a plurality of pictures to be identified;
acquiring a plurality of target pictures from the plurality of pictures to be identified based on a house indoor scene classification model;
the step of executing the flip recognition processing on the target picture comprises the following steps: respectively executing a roll-over identification process on each of the plurality of target pictures;
the method further comprises the steps of: and under the condition that the occupation ratio of the first picture in the plurality of target pictures is larger than a proportion threshold value, determining that the video is a flip video, wherein the first picture is the target picture determined to be the flip picture.
7. The method of claim 6, wherein the taking a plurality of video frames in the video to be identified as a plurality of pictures to be identified comprises:
and performing frame extraction processing on the video to be identified to obtain the plurality of video frames.
8. An apparatus for identifying a flip image, the apparatus comprising:
the acquisition module is used for acquiring a target picture based on the house indoor scene classification model, wherein the target picture is a house indoor scene picture;
the device comprises a reproduction identification processing module, a reproduction identification processing module and a reproduction identification processing module, wherein the reproduction identification processing module is used for executing reproduction identification processing on the target picture, and the reproduction identification processing comprises: and determining that the target picture is a flip picture when the flip probability of the target picture is larger than a probability threshold value or the picture identification result of the target picture indicates that a flip mark exists.
9. An electronic device, comprising: a processor, a memory and a computer program stored on the memory and executable on the processor, which when executed by the processor performs the steps of the method of identifying a flip image as claimed in any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which computer program, when being executed by a processor, implements the steps of the method for identifying a flip image as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310799255.8A CN116958795A (en) | 2023-06-30 | 2023-06-30 | Method and device for identifying flip image, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310799255.8A CN116958795A (en) | 2023-06-30 | 2023-06-30 | Method and device for identifying flip image, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116958795A true CN116958795A (en) | 2023-10-27 |
Family
ID=88443630
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310799255.8A Pending CN116958795A (en) | 2023-06-30 | 2023-06-30 | Method and device for identifying flip image, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116958795A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117333762A (en) * | 2023-12-02 | 2024-01-02 | 深圳爱莫科技有限公司 | Image reproduction identification method based on multi-feature fusion |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107729847A (en) * | 2017-10-20 | 2018-02-23 | 阿里巴巴集团控股有限公司 | A kind of certificate verification, auth method and device |
| CN109583909A (en) * | 2018-10-26 | 2019-04-05 | 深圳壹账通智能科技有限公司 | Houseclearing discriminating method, device, equipment and medium based on image recognition |
| CN109816004A (en) * | 2019-01-17 | 2019-05-28 | 平安城市建设科技(深圳)有限公司 | Source of houses picture classification method, device, equipment and storage medium |
| CN110659604A (en) * | 2019-09-20 | 2020-01-07 | 北京达佳互联信息技术有限公司 | Video detection method, device, server and storage medium |
| CN111461143A (en) * | 2020-03-31 | 2020-07-28 | 珠海格力电器股份有限公司 | Picture copying identification method and device and electronic equipment |
| CN112257709A (en) * | 2020-10-23 | 2021-01-22 | 北京云杉世界信息技术有限公司 | Signboard photo auditing method and device, electronic equipment and readable storage medium |
| CN112650875A (en) * | 2020-12-22 | 2021-04-13 | 深圳壹账通智能科技有限公司 | House image verification method and device, computer equipment and storage medium |
| CN113111888A (en) * | 2021-04-15 | 2021-07-13 | 广州图匠数据科技有限公司 | Picture distinguishing method and device |
| CN113222952A (en) * | 2021-05-20 | 2021-08-06 | 支付宝(杭州)信息技术有限公司 | Method and device for identifying copied image |
| CN114782797A (en) * | 2022-06-21 | 2022-07-22 | 深圳市万物云科技有限公司 | House scene classification method, device and equipment and readable storage medium |
| WO2023071609A1 (en) * | 2021-10-29 | 2023-05-04 | 北京有竹居网络技术有限公司 | Copied image recognition method and related device thereof |
| CN116168038A (en) * | 2023-04-26 | 2023-05-26 | 创新奇智(青岛)科技有限公司 | Image reproduction detection method and device, electronic equipment and storage medium |
-
2023
- 2023-06-30 CN CN202310799255.8A patent/CN116958795A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107729847A (en) * | 2017-10-20 | 2018-02-23 | 阿里巴巴集团控股有限公司 | A kind of certificate verification, auth method and device |
| CN109583909A (en) * | 2018-10-26 | 2019-04-05 | 深圳壹账通智能科技有限公司 | Houseclearing discriminating method, device, equipment and medium based on image recognition |
| CN109816004A (en) * | 2019-01-17 | 2019-05-28 | 平安城市建设科技(深圳)有限公司 | Source of houses picture classification method, device, equipment and storage medium |
| CN110659604A (en) * | 2019-09-20 | 2020-01-07 | 北京达佳互联信息技术有限公司 | Video detection method, device, server and storage medium |
| CN111461143A (en) * | 2020-03-31 | 2020-07-28 | 珠海格力电器股份有限公司 | Picture copying identification method and device and electronic equipment |
| CN112257709A (en) * | 2020-10-23 | 2021-01-22 | 北京云杉世界信息技术有限公司 | Signboard photo auditing method and device, electronic equipment and readable storage medium |
| CN112650875A (en) * | 2020-12-22 | 2021-04-13 | 深圳壹账通智能科技有限公司 | House image verification method and device, computer equipment and storage medium |
| CN113111888A (en) * | 2021-04-15 | 2021-07-13 | 广州图匠数据科技有限公司 | Picture distinguishing method and device |
| CN113222952A (en) * | 2021-05-20 | 2021-08-06 | 支付宝(杭州)信息技术有限公司 | Method and device for identifying copied image |
| WO2023071609A1 (en) * | 2021-10-29 | 2023-05-04 | 北京有竹居网络技术有限公司 | Copied image recognition method and related device thereof |
| CN114782797A (en) * | 2022-06-21 | 2022-07-22 | 深圳市万物云科技有限公司 | House scene classification method, device and equipment and readable storage medium |
| CN116168038A (en) * | 2023-04-26 | 2023-05-26 | 创新奇智(青岛)科技有限公司 | Image reproduction detection method and device, electronic equipment and storage medium |
Non-Patent Citations (2)
| Title |
|---|
| ELDHO ABRAHAM等: "Moir´e Pattern Detection using Wavelet Decomposition and Convolutional Neural Network", 《SSCI》, pages 1275 - 1279 * |
| 汪振杰: "基于深度学习的货架商品目标检测", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 2020, pages 138 - 640 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117333762A (en) * | 2023-12-02 | 2024-01-02 | 深圳爱莫科技有限公司 | Image reproduction identification method based on multi-feature fusion |
| CN117333762B (en) * | 2023-12-02 | 2024-03-15 | 深圳爱莫科技有限公司 | Image reproduction identification method based on multi-feature fusion |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109325933B (en) | Method and device for recognizing copied image | |
| EP3916627A1 (en) | Living body detection method based on facial recognition, and electronic device and storage medium | |
| CN106326832B (en) | Device and method for processing image based on object region | |
| US11176355B2 (en) | Facial image processing method and apparatus, electronic device and computer readable storage medium | |
| CN108230333B (en) | Image processing method, image processing apparatus, computer program, storage medium, and electronic device | |
| CN110807759A (en) | Method and device for evaluating photo quality, electronic equipment and readable storage medium | |
| WO2022160857A1 (en) | Image processing method and apparatus, and computer-readable storage medium and electronic device | |
| WO2021008205A1 (en) | Image processing | |
| AU2020309094B2 (en) | Image processing method and apparatus, electronic device, and storage medium | |
| JP5937823B2 (en) | Image collation processing apparatus, image collation processing method, and image collation processing program | |
| CN116958795A (en) | Method and device for identifying flip image, electronic equipment and storage medium | |
| CN111325107A (en) | Detection model training method and device, electronic equipment and readable storage medium | |
| CN112802081A (en) | Depth detection method and device, electronic equipment and storage medium | |
| CN109726613B (en) | Method and device for detection | |
| JP5651659B2 (en) | Object detection system and program | |
| CN106611417B (en) | Method and device for classifying visual elements into foreground or background | |
| CN107578006B (en) | Photo processing method and mobile terminal | |
| CN106402717B (en) | A kind of AR control method for playing back and intelligent desk lamp | |
| CN111767845B (en) | Certificate identification method and device | |
| CN112818743A (en) | Image recognition method and device, electronic equipment and computer storage medium | |
| CN111126283A (en) | Rapid in-vivo detection method and system for automatically filtering fuzzy human face | |
| CN113691731B (en) | Processing method and device and electronic equipment | |
| CN114143442B (en) | Image blurring method, computer device, and computer-readable storage medium | |
| CN110849317B (en) | Method for determining included angle between display screens, electronic equipment and storage medium | |
| Li et al. | Image sharpness evaluation based on visual importance |
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 |