CN114387658A - Image target attribute detection method, device, equipment and storage medium - Google Patents
Image target attribute detection method, device, equipment and storage medium Download PDFInfo
- Publication number
- CN114387658A CN114387658A CN202210292392.8A CN202210292392A CN114387658A CN 114387658 A CN114387658 A CN 114387658A CN 202210292392 A CN202210292392 A CN 202210292392A CN 114387658 A CN114387658 A CN 114387658A
- Authority
- CN
- China
- Prior art keywords
- target
- image
- deep learning
- learning model
- trained
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Image Analysis (AREA)
Abstract
The application discloses a method, a device, equipment and a storage medium for detecting image target attributes, which relate to the field of computer vision and comprise the following steps: training a first deep learning model by utilizing a preset target frame training set to obtain a trained first deep learning model, detecting an image to be detected to obtain a first detection result, and cutting the image to be detected according to the first detection result to obtain a corresponding target subgraph; training a second deep learning model based on a preset key point training set to obtain a trained second deep learning model, and detecting the target subgraph to obtain a second detection result; and acquiring corresponding target information according to the second detection result, and detecting the target attribute of the image to be detected based on the target information. According to the method and the device, the information in the image to be detected is fully utilized by combining the target frame detection and the key point detection, and the accuracy of target attribute detection is effectively improved.
Description
Technical Field
The invention relates to the field of computer vision, in particular to a method, a device, equipment and a storage medium for detecting image target attributes.
Background
Currently, Object Detection (Object Detection) is one of the basic tasks in the field of computer vision, and in the computer vision task, generally, the Detection of the Object attribute is performed after detecting the Object based on an Object Detection model, but the Detection of the attribute based on the Detection result of the Object Detection model only causes the problem that the information in the image is not sufficiently utilized, for example, when the Object in the image is detected by the Object Detection model, the attitude information and the motion information of the Object cannot be acquired, which causes the problem that the accuracy of the Detection of the Object attribute of the image is low.
Disclosure of Invention
In view of the above, the present invention provides a method, an apparatus, a device and a storage medium for detecting an image target attribute, which can effectively solve the problem of target attribute classification in a computer vision task and can also effectively improve the accuracy of image target attribute detection. The specific scheme is as follows:
in a first aspect, the present application discloses a method for detecting an image target attribute, including:
training a first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and training a second deep learning model based on a preset key point training set to obtain a trained second deep learning model;
detecting an image to be detected by using the trained first deep learning model to obtain a first detection result, and cutting the image to be detected according to the first detection result to obtain a corresponding target subgraph;
detecting the target subgraph by using the trained second deep learning model to obtain a second detection result;
and acquiring corresponding target information according to the second detection result, and detecting the target attribute of the image to be detected based on the target information.
Optionally, before the training of the first deep learning model based on the preset target frame training set to obtain the trained first deep learning model, the method further includes:
determining a target frame and key points in an image to be detected in an image data set, and marking the target frame and the key points to obtain a preset target frame training set corresponding to the target frame and a preset key point training set corresponding to the key points.
Optionally, the training of the first deep learning model based on the preset target frame training set to obtain a trained first deep learning model, and the training of the second deep learning model based on the preset key point training set to obtain a trained second deep learning model include:
and training a first deep learning model constructed based on a first deep learning algorithm by using the preset target frame training set to obtain a trained first deep learning model, and training a second deep learning model constructed based on a second deep learning algorithm by using the preset key point training set to obtain a trained second deep learning model.
Optionally, the detecting the image to be detected by using the trained first deep learning model to obtain a first detection result includes:
performing target detection on an image to be detected based on the trained first deep learning model to obtain corresponding target frame coordinate information, and cutting the image to be detected based on the target frame coordinate information to obtain a corresponding target sub-image;
correspondingly, the detecting the target sub-graph by using the trained second deep learning model to obtain a second detection result includes:
and performing key point detection on the target subgraph based on the trained second deep learning model to obtain corresponding key point coordinate information.
Optionally, the performing, based on the trained first deep learning model, target detection on the image to be detected to obtain corresponding target frame coordinate information, and clipping the image to be detected to obtain a corresponding target sub-image based on the target frame coordinate information includes:
carrying out pedestrian detection on an image to be detected based on the trained first deep learning model to obtain corresponding pedestrian frame coordinate information, and cutting the image to be detected based on the pedestrian frame coordinate information to obtain a corresponding pedestrian sub-image;
and/or the presence of a gas in the gas,
and carrying out face detection on the image to be detected based on the trained first deep learning model to obtain corresponding face frame coordinate information, and cutting the image to be detected based on the face frame coordinate information to obtain a corresponding face subgraph.
Optionally, the performing, based on the trained second deep learning model, key point detection on the target sub-graph to obtain corresponding key point coordinate information includes:
carrying out key point detection on the pedestrian subgraph based on the trained second deep learning model to obtain corresponding human body key point coordinate information;
and/or the presence of a gas in the gas,
and performing key point detection on the face subgraph based on the trained second deep learning model to obtain corresponding face key point coordinate information.
Optionally, the obtaining of the corresponding target information according to the second detection result includes:
acquiring target image information and key point information of the image to be detected based on the second detection result;
correspondingly, the detecting the target attribute of the target to be detected based on the target information includes:
and detecting the target attribute of the image to be detected based on the target image information and the key point information.
In a second aspect, the present application discloses an image target property detection apparatus, comprising:
the model training module is used for training the first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and training the second deep learning model based on a preset key point training set to obtain a trained second deep learning model;
the first image detection module is used for detecting an image to be detected by using the trained first deep learning model to obtain a first detection result, and cutting the image to be detected according to the first detection result to obtain a corresponding target subgraph;
the second image detection module is used for detecting the target subgraph by using the trained second deep learning model to obtain a second detection result;
the information acquisition module is used for acquiring corresponding target information according to the second detection result;
and the target attribute detection module is used for detecting the target attribute of the image to be detected based on the target information.
In a third aspect, the present application discloses an image target attribute detection apparatus, comprising:
a memory for storing a computer program;
a processor for executing the computer program to implement the steps of the image target property detection method disclosed in the foregoing.
In a fourth aspect, the present application discloses a computer readable storage medium for storing a computer program; wherein the computer program realizes the steps of the image object property detection method disclosed in the foregoing when being executed by a processor.
Therefore, the application provides an image target attribute detection method, which comprises the following steps: training a first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and training a second deep learning model based on a preset key point training set to obtain a trained second deep learning model; detecting an image to be detected by using the trained first deep learning model to obtain a first detection result, and cutting the image to be detected according to the first detection result to obtain a corresponding target subgraph; detecting the target subgraph by using the trained second deep learning model to obtain a second detection result; and acquiring corresponding target information according to the second detection result, and detecting the target attribute of the image to be detected based on the target information. Therefore, the target detection is carried out on the image to be detected by utilizing the first deep learning model, then the key point detection is carried out on the target subgraph obtained based on the first detection result by utilizing the second deep learning model, and finally corresponding target information is obtained according to the second detection result, so that the target attribute of the image to be detected is detected and analyzed based on the target information. According to the method and the device for detecting the target attribute, the mode of combining the target detection and the key point detection is carried out based on the deep learning model, information in the image to be detected is fully utilized, the accuracy of target attribute detection is effectively improved, and the problem of target attribute classification in a computer vision task can be effectively solved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the provided drawings without creative efforts.
FIG. 1 is a flowchart of an image target attribute detection method disclosed in the present application;
FIG. 2 is a flowchart of a specific image target attribute detection method disclosed in the present application;
FIG. 3 is a schematic structural diagram of an image target attribute detection apparatus disclosed in the present application;
fig. 4 is a structural diagram of an image target attribute detection device disclosed in the present application.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Currently, in a computer vision task, detecting target attributes generally includes detecting targets based on a target detection model and then detecting the target attributes, but performing attribute detection based on only a detection result of the target detection model may cause a problem of insufficient utilization of information in an image, resulting in low accuracy of image target attribute detection. Therefore, the image target attribute detection scheme is provided, the problem of target attribute classification in a computer vision task can be effectively solved, and the accuracy of image target attribute detection can be effectively improved.
The embodiment of the invention discloses an image target attribute detection method, which is shown in figure 1 and comprises the following steps:
step S11: and training the first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and training the second deep learning model based on a preset key point training set to obtain a trained second deep learning model.
In this embodiment, before training the deep learning model, a related training set needs to be determined, and then the deep learning model is trained by using the training set. Specifically, a target frame and key points in an image to be detected in an image data set are determined, and the target frame and the key points are marked to obtain a preset target frame training set corresponding to the target frame and a preset key point training set corresponding to the key points. It will be appreciated that in marking a target border in the image dataset, the marking of the target border and the keypoints may include, but is not limited to. The target frame may further include, but is not limited to, a pedestrian target frame or a human face target frame. When the above key points are marked, the key points may include, but are not limited to, key points of a pedestrian, such as joints of the pedestrian, or key points of a human face, such as corners of eyes or mouths of the human face. The data are marked to obtain a corresponding training set, namely, some key position information is provided for the deep learning model, so that the deep learning model can accurately obtain the position information of the target frame and the key points.
In this embodiment, after the preset target frame training set and the preset key point training set are obtained, the first deep learning model is trained based on the preset target frame training set to obtain a trained first deep learning model, and the second deep learning model is trained based on the preset key point training set to obtain a trained second deep learning model. Namely, the preset target frame training set is used for training a first deep learning model constructed based on a first deep learning algorithm to obtain a trained first deep learning model, and the preset key point training set is used for training a second deep learning model constructed based on a second deep learning algorithm to obtain a trained second deep learning model. It can be understood that the first deep learning model and the second deep learning model are blank models, the blank models are trained by using the preset target frame training set and the preset key point training set, and then the trained first deep learning model and the trained second deep learning model can be obtained, the trained first deep learning model can include but is not limited to a target detection model and a face detection model, and the trained second deep learning model can include but is not limited to a human key point detection model and a face key point detection model.
Step S12: and detecting an image to be detected by using the trained first deep learning model to obtain a first detection result, and cutting the image to be detected according to the first detection result to obtain a corresponding target subgraph.
In this embodiment, obtaining the trained first deep learning model and the trained second deep learning model can detect an image to be detected by using the trained first deep learning model to obtain a first detection result, and cut the image to be detected according to the first detection result to obtain a corresponding target sub-image. It can be understood that the target in the image to be detected is detected based on the trained first deep learning model to obtain a first detection result, that is, the image to be detected is input to the trained first deep learning model to obtain frame information of the target corresponding to the image to be detected and output by the trained first deep learning model, and the image to be detected is clipped based on the frame information of the target, so as to obtain a desired target sub-graph.
Step S13: and detecting the target subgraph by using the trained second deep learning model to obtain a second detection result.
In this embodiment, after the image to be detected is cut according to the first detection result to obtain a corresponding target sub-image, the trained second deep learning model is used to detect the target sub-image to obtain a second detection result. Namely, the target sub-graph is input into the trained second deep learning model, so as to obtain the target key point information which is output by the trained second deep learning model and corresponds to the target sub-graph. It can be understood that, the target subgraph is subjected to key point detection based on the trained second deep learning model so as to acquire the position of the relevant key point in the target based on the first detection, so that all information in the image to be detected, such as the joint position of a person or the position information of the corner of the mouth and the corner of the eye of a face, can be effectively utilized.
Step S14: and acquiring corresponding target information according to the second detection result, and detecting the target attribute of the image to be detected based on the target information.
In this embodiment, since the second detection result includes key position information of the target, target image information and target key point information of the image to be detected can be obtained based on the second detection result; and detecting the target attribute of the image to be detected based on the target image information and the target key point information. It can be understood that the target information in the image to be detected can be accurately obtained based on the position information of the target key point detected by the key point, and the information such as the posture and the motion of the target can also be obtained through the position information of the target key point, so that the target attribute can be detected by inputting the target image information and the posture and motion information into the attribute analysis model or performing attribute analysis on the target image information and the posture and motion information by using a corresponding attribute analysis algorithm, thereby obtaining a corresponding attribute analysis result.
It should be noted that, after the target data in the image data set is marked to obtain the corresponding marked data, the target attribute can also be directly analyzed based on the marked data, so that the information in the image to be detected can be fully utilized, and the accuracy of target attribute detection can be improved. Namely, determining a target frame and key points in an image to be detected in an image data set, marking the target frame and the key points to obtain a preset target frame training set corresponding to the target frame and a preset key point training set corresponding to the key points, and then directly analyzing target attributes by using the target frame and the key points marked in the preset target frame training set and the preset key point training set.
Therefore, in the embodiment of the application, the first deep learning model is used for carrying out target detection on the image to be detected, then the second deep learning model is used for carrying out key point detection on the target subgraph obtained based on the first detection result, and finally corresponding target information is obtained according to the second detection result, so that the target attribute of the image to be detected is detected and analyzed based on the target information. The method and the device for detecting the target attribute of the computer vision task have the advantages that the mode of combining the target detection and the key point detection based on the deep learning model fully utilizes the information in the image to be detected, the accuracy of the target attribute detection is effectively improved, and the problem of target attribute classification in the computer vision task can be effectively solved.
Referring to fig. 2, the embodiment of the present invention discloses a specific image target attribute detection method, and compared with the previous embodiment, the present embodiment further describes and optimizes the technical solution.
Step S21: and training the first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and training the second deep learning model based on a preset key point training set to obtain a trained second deep learning model.
Step S22: and performing target detection on the image to be detected based on the trained first deep learning model to obtain corresponding target frame coordinate information, and cutting the image to be detected based on the target frame coordinate information to obtain a corresponding target sub-image.
In this embodiment, the target detection is performed on the image to be detected based on the trained first deep learning model to obtain corresponding target frame coordinate information. It can be understood that, by using the trained first deep learning model to detect the target in the image to be detected, the image coordinates of the frame of the target can be obtained: (x1, y1), (x1, y2), (x2, y1), (x2, y2), so that the image to be detected can be cropped based on the image coordinates to obtain a target subgraph.
Step S23: and performing key point detection on the target subgraph based on the trained second deep learning model to obtain corresponding key point coordinate information.
In this embodiment, the key point detection is performed on the target sub-graph based on the trained second deep learning model to obtain corresponding key point coordinate information. It can be understood that, the trained second deep learning model is used to perform key point detection on the target sub-graph, and image coordinates of the target key points are obtained: (x1, y1), (x2, y2), …, (xn, yn), so that information such as the posture and the like of the object can be acquired by using the image coordinates of the object key point when the object attribute analysis is performed.
In a specific embodiment, the performing, based on the trained first deep learning model, target detection on an image to be detected to obtain corresponding target frame coordinate information, and cropping the image to be detected to obtain a corresponding target sub-image based on the target frame coordinate information may include: and carrying out pedestrian detection on the image to be detected based on the trained first deep learning model to obtain corresponding pedestrian frame coordinate information, and cutting the image to be detected based on the pedestrian frame coordinate information to obtain a corresponding pedestrian sub-image. Correspondingly, carrying out key point detection on the pedestrian subgraph based on the trained second deep learning model to obtain corresponding human body key point coordinate information. It can be understood that, when the trained first deep learning model is a pedestrian target detection model, an image to be detected is input into the target detection model, and the target detection model outputs pedestrian frame coordinate information corresponding to the image to be detected, so that a corresponding pedestrian sub-graph is cut out based on the pedestrian frame information. And then inputting the pedestrian subgraph into the human body key point detection model for detection so as to obtain the coordinate information of the human body key point.
In a specific embodiment, the performing, based on the trained first deep learning model, target detection on an image to be detected to obtain corresponding target frame coordinate information, and cropping the image to be detected to obtain a corresponding target sub-image based on the target frame coordinate information may include: and carrying out face detection on the image to be detected based on the trained first deep learning model to obtain corresponding face frame coordinate information, and cutting the image to be detected based on the face frame coordinate information to obtain a corresponding face subgraph. Correspondingly, carrying out key point detection on the face subgraph based on the trained second deep learning model to obtain corresponding face key point coordinate information. It can be understood that, when the trained first deep learning model is a face detection model, an image to be detected is input into the face detection model, the face detection model outputs face frame coordinate information corresponding to the image to be detected, a corresponding face sub-image is cut out based on the face frame information, and then the face sub-image is input into the face key point detection model for detection, so that face key point coordinate information is obtained.
Step S24: and acquiring corresponding target information according to the coordinate information of the key points, and detecting the target attribute of the image to be detected based on the target information.
For the specific contents of the step S21 and the step S24, reference may be made to the corresponding contents disclosed in the foregoing embodiments, and details are not repeated here.
Correspondingly, the embodiment of the present application further discloses an image target attribute detection apparatus, as shown in fig. 3, the apparatus includes:
the model training module 11 is configured to train a first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and train a second deep learning model based on a preset key point training set to obtain a trained second deep learning model;
the first image detection module 12 is configured to detect an image to be detected by using the trained first deep learning model to obtain a first detection result, and cut the image to be detected according to the first detection result to obtain a corresponding target sub-image;
the second image detection module 13 is configured to detect the target sub-image by using the trained second deep learning model to obtain a second detection result;
the information acquisition module 14 is configured to acquire corresponding target information according to the second detection result;
and the target attribute detection module 15 is configured to detect a target attribute of the image to be detected based on the target information.
As can be seen from the above, in this embodiment, the first deep learning model is used to perform target detection on an image to be detected, then the second deep learning model is used to perform key point detection on a target sub-image obtained based on the first detection result, and finally corresponding target information is obtained according to the second detection result, so that the target attribute of the image to be detected is detected and analyzed based on the target information. The method and the device for detecting the target attribute of the computer vision task have the advantages that the mode of combining the target detection and the key point detection based on the deep learning model fully utilizes the information in the image to be detected, the accuracy of the target attribute detection is effectively improved, and the problem of target attribute classification in the computer vision task can be effectively solved.
In some specific embodiments, the image target property detection apparatus further includes:
and the marking module is used for determining a target frame and key points in the image to be detected in the image data set and marking the target frame and the key points so as to obtain a preset target frame training set corresponding to the target frame and a preset key point training set corresponding to the key points.
In some specific embodiments, the model training module 11 specifically includes:
the first module training unit is used for training a first deep learning model constructed based on a first deep learning algorithm by using the preset target frame training set to obtain a trained first deep learning model;
and the second module training unit is used for training a second deep learning model constructed based on a second deep learning algorithm by using the preset key point training set to obtain the trained second deep learning model.
In some specific embodiments, the first image detection module 12 specifically includes:
the target detection unit is used for carrying out target detection on the image to be detected based on the trained first deep learning model to obtain corresponding target frame coordinate information and cutting the image to be detected based on the target frame coordinate information to obtain a corresponding target sub-image;
in some specific embodiments, the second image detection module 13 specifically includes:
and the key point detection unit is used for detecting key points of the target subgraph based on the trained second deep learning model to obtain corresponding key point coordinate information.
In some specific embodiments, the first image detection module 12 specifically includes:
the first target detection unit is used for carrying out pedestrian detection on the image to be detected based on the trained first deep learning model to obtain corresponding pedestrian frame coordinate information and cutting the image to be detected based on the pedestrian frame coordinate information to obtain a corresponding pedestrian sub-image;
and the second target detection unit is used for carrying out face detection on the image to be detected based on the trained first deep learning model to obtain corresponding face frame coordinate information and cutting the image to be detected based on the line face frame coordinate information to obtain a corresponding face subgraph.
In some specific embodiments, the second image detection module 13 specifically includes:
the first key point detection unit is used for detecting key points of the pedestrian subgraph on the basis of the trained second deep learning model to obtain corresponding human body key point coordinate information;
and the second key point detection unit is used for carrying out key point detection on the face subgraph based on the trained second deep learning model so as to obtain corresponding face key point coordinate information.
In some specific embodiments, the information obtaining module 14 specifically includes:
the information acquisition unit is used for acquiring target image information and key point information of the image to be detected based on the second detection result;
in some specific embodiments, the target attribute detection module 15 specifically includes:
and the target attribute detection unit is used for detecting the target attribute of the image to be detected based on the target image information and the key point information.
Further, the embodiment of the application also provides an image target attribute detection device. FIG. 4 is a block diagram illustrating an image target property detection device 20 according to an exemplary embodiment, and the contents of the diagram should not be construed as limiting the scope of use of the present application in any way.
Fig. 4 is a schematic structural diagram of an image target attribute detection device 20 according to an embodiment of the present application. The image target property detection device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input output interface 25, and a communication bus 26. The memory 22 is used for storing a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the image target property detection method disclosed in any of the foregoing embodiments. In addition, the image target attribute detection device 20 in the present embodiment may specifically be an electronic computer.
In this embodiment, the power supply 23 is configured to provide a working voltage for each hardware device on the image target attribute detection device 20; the communication interface 24 can create a data transmission channel between the image target property detection device 20 and an external device, and a communication protocol followed by the communication interface is any communication protocol applicable to the technical solution of the present application, and is not specifically limited herein; the input/output interface 25 is configured to obtain external input data or output data to the outside, and a specific interface type thereof may be selected according to specific application requirements, which is not specifically limited herein.
In addition, the storage 22 is used as a carrier for resource storage, and may be a read-only memory, a random access memory, a magnetic disk or an optical disk, etc., and the resources stored thereon may include an operating system 221, a computer program 222, etc., and the storage manner may be a transient storage or a permanent storage.
The operating system 221 is used to manage and control each hardware device on the image target attribute detection device 20 and the computer program 222, and may be Windows Server, Netware, Unix, Linux, or the like. The computer program 222 may further include a computer program that can be used to perform other specific tasks in addition to the computer program that can be used to perform the image target property detection method performed by the image target property detection apparatus 20 disclosed in any of the foregoing embodiments.
Further, an embodiment of the present application further discloses a storage medium, in which a computer program is stored, and when the computer program is loaded and executed by a processor, the steps of the image target attribute detection method disclosed in any of the foregoing embodiments are implemented.
The embodiments are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same or similar parts among the embodiments are referred to each other. The device disclosed by the embodiment corresponds to the method disclosed by the embodiment, so that the description is simple, and the relevant points can be referred to the method part for description.
Finally, it should also be noted that, herein, relational terms such as first and second, and the like may be 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. Also, 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 an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The method, the apparatus, the device and the storage medium for detecting the image target attribute provided by the present invention are described in detail above, and a specific example is applied in the present document to illustrate the principle and the implementation manner of the present invention, and the description of the above embodiment is only used to help understanding the method and the core idea of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, there may be variations in the specific embodiments and the application scope, and in summary, the content of the present specification should not be construed as a limitation to the present invention.
Claims (10)
1. An image target attribute detection method is characterized by comprising the following steps:
training a first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and training a second deep learning model based on a preset key point training set to obtain a trained second deep learning model;
detecting an image to be detected by using the trained first deep learning model to obtain a first detection result, and cutting the image to be detected according to the first detection result to obtain a corresponding target subgraph;
detecting the target subgraph by using the trained second deep learning model to obtain a second detection result;
and acquiring corresponding target information according to the second detection result, and detecting the target attribute of the image to be detected based on the target information.
2. The image target attribute detection method according to claim 1, wherein before the training of the first deep learning model based on the preset target frame training set to obtain the trained first deep learning model, the method further comprises:
determining a target frame and key points in an image to be detected in an image data set, and marking the target frame and the key points to obtain a preset target frame training set corresponding to the target frame and a preset key point training set corresponding to the key points.
3. The image target attribute detection method according to claim 2, wherein the training of the first deep learning model based on the preset target frame training set to obtain the trained first deep learning model and the training of the second deep learning model based on the preset key point training set to obtain the trained second deep learning model comprises:
and training a first deep learning model constructed based on a first deep learning algorithm by using the preset target frame training set to obtain a trained first deep learning model, and training a second deep learning model constructed based on a second deep learning algorithm by using the preset key point training set to obtain a trained second deep learning model.
4. The image target property detection method according to claim 1, wherein the detecting an image to be detected by using the trained first deep learning model to obtain a first detection result includes:
performing target detection on an image to be detected based on the trained first deep learning model to obtain corresponding target frame coordinate information, and cutting the image to be detected based on the target frame coordinate information to obtain a corresponding target sub-image;
correspondingly, the detecting the target sub-graph by using the trained second deep learning model to obtain a second detection result includes:
and performing key point detection on the target subgraph based on the trained second deep learning model to obtain corresponding key point coordinate information.
5. The image target attribute detection method according to claim 4, wherein the performing target detection on the image to be detected based on the trained first deep learning model to obtain corresponding target frame coordinate information, and cropping the image to be detected based on the target frame coordinate information to obtain a corresponding target sub-graph comprises:
carrying out pedestrian detection on an image to be detected based on the trained first deep learning model to obtain corresponding pedestrian frame coordinate information, and cutting the image to be detected based on the pedestrian frame coordinate information to obtain a corresponding pedestrian sub-image;
and/or the presence of a gas in the gas,
and carrying out face detection on the image to be detected based on the trained first deep learning model to obtain corresponding face frame coordinate information, and cutting the image to be detected based on the face frame coordinate information to obtain a corresponding face subgraph.
6. The image target property detection method according to claim 5, wherein the performing key point detection on the target sub-graph based on the trained second deep learning model to obtain corresponding key point coordinate information comprises:
carrying out key point detection on the pedestrian subgraph based on the trained second deep learning model to obtain corresponding human body key point coordinate information;
and/or the presence of a gas in the gas,
and performing key point detection on the face subgraph based on the trained second deep learning model to obtain corresponding face key point coordinate information.
7. The image target property detection method according to any one of claims 1 to 6, wherein the obtaining of the corresponding target information according to the second detection result includes:
acquiring target image information and key point information of the image to be detected based on the second detection result;
correspondingly, the detecting the target attribute of the target to be detected based on the target information includes:
and detecting the target attribute of the image to be detected based on the target image information and the key point information.
8. An image object property detection apparatus, comprising:
the model training module is used for training the first deep learning model based on a preset target frame training set to obtain a trained first deep learning model, and training the second deep learning model based on a preset key point training set to obtain a trained second deep learning model;
the first image detection module is used for detecting an image to be detected by using the trained first deep learning model to obtain a first detection result, and cutting the image to be detected according to the first detection result to obtain a corresponding target subgraph;
the second image detection module is used for detecting the target subgraph by using the trained second deep learning model to obtain a second detection result;
the information acquisition module is used for acquiring corresponding target information according to the second detection result;
and the target attribute detection module is used for detecting the target attribute of the image to be detected based on the target information.
9. An image object property detection apparatus characterized by comprising:
a memory for storing a computer program;
a processor for executing the computer program for carrying out the steps of the image target property detection method according to any one of claims 1 to 7.
10. A computer-readable storage medium for storing a computer program; wherein the computer program realizes the steps of the image object property detection method according to any one of claims 1 to 7 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210292392.8A CN114387658A (en) | 2022-03-24 | 2022-03-24 | Image target attribute detection method, device, equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210292392.8A CN114387658A (en) | 2022-03-24 | 2022-03-24 | Image target attribute detection method, device, equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114387658A true CN114387658A (en) | 2022-04-22 |
Family
ID=81205732
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210292392.8A Pending CN114387658A (en) | 2022-03-24 | 2022-03-24 | Image target attribute detection method, device, equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114387658A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110287856A (en) * | 2019-06-21 | 2019-09-27 | 上海闪马智能科技有限公司 | A kind of security personnel's behavior analysis system, method and device |
| CN115035294A (en) * | 2022-06-15 | 2022-09-09 | 灵图数据(杭州)有限公司 | Refrigerator temperature controller gear detection method and device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558864A (en) * | 2019-01-16 | 2019-04-02 | 苏州科达科技股份有限公司 | Face critical point detection method, apparatus and storage medium |
| CN112819885A (en) * | 2021-02-20 | 2021-05-18 | 深圳市英威诺科技有限公司 | Animal identification method, device and equipment based on deep learning and storage medium |
| CN113505682A (en) * | 2021-07-02 | 2021-10-15 | 杭州萤石软件有限公司 | Living body detection method and device |
-
2022
- 2022-03-24 CN CN202210292392.8A patent/CN114387658A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109558864A (en) * | 2019-01-16 | 2019-04-02 | 苏州科达科技股份有限公司 | Face critical point detection method, apparatus and storage medium |
| CN112819885A (en) * | 2021-02-20 | 2021-05-18 | 深圳市英威诺科技有限公司 | Animal identification method, device and equipment based on deep learning and storage medium |
| CN113505682A (en) * | 2021-07-02 | 2021-10-15 | 杭州萤石软件有限公司 | Living body detection method and device |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110287856A (en) * | 2019-06-21 | 2019-09-27 | 上海闪马智能科技有限公司 | A kind of security personnel's behavior analysis system, method and device |
| CN115035294A (en) * | 2022-06-15 | 2022-09-09 | 灵图数据(杭州)有限公司 | Refrigerator temperature controller gear detection method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6588053B2 (en) | Method and apparatus for annotating point cloud data | |
| CN114387658A (en) | Image target attribute detection method, device, equipment and storage medium | |
| CN109582880B (en) | Interest point information processing method, device, terminal and storage medium | |
| CN111708366B (en) | Robot, and method, apparatus and computer-readable storage medium for controlling movement of robot | |
| CN111027403B (en) | Gesture estimation method, device, equipment and computer readable storage medium | |
| CN110390261B (en) | Target detection method and device, computer readable storage medium and electronic equipment | |
| CN110969045A (en) | Behavior detection method and device, electronic equipment and storage medium | |
| US11798181B2 (en) | Method and system for location detection of photographs using topographic techniques | |
| CN111881740A (en) | Face recognition method, face recognition device, electronic equipment and medium | |
| CN113516697B (en) | Image registration method, device, electronic equipment and computer readable storage medium | |
| CN105302576B (en) | A kind of users' guidebook generation method and system updated based on APP | |
| CN113591885A (en) | Target detection model training method, device and computer storage medium | |
| CN111027376A (en) | Method and device for determining event map, electronic equipment and storage medium | |
| CN113850239B (en) | Multi-document detection method and device, electronic equipment and storage medium | |
| CN110084187A (en) | Location recognition method, device, equipment and storage medium based on computer vision | |
| CN113063424B (en) | Method, device, equipment and storage medium for intra-market navigation | |
| CN112860060B (en) | Image recognition method, device and storage medium | |
| CN114429631A (en) | Three-dimensional object detection method, device, equipment and storage medium | |
| US20240070516A1 (en) | Machine learning context based confidence calibration | |
| CN116450917B (en) | Information searching method and device, electronic equipment and medium | |
| CN113469878B (en) | Text erasing method and training method and device of model thereof, and storage medium | |
| CN113592981B (en) | Picture labeling method and device, electronic equipment and storage medium | |
| CN110717576B (en) | Image processing method, device and equipment | |
| CN115273102A (en) | Method, device, equipment and medium for grading handwritten text neatness | |
| CN117152585A (en) | Video action recognition method, device, equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220422 |