CN111126274A - Method, device, equipment and medium for detecting inbound target population - Google Patents
Method, device, equipment and medium for detecting inbound target population Download PDFInfo
- Publication number
- CN111126274A CN111126274A CN201911348895.7A CN201911348895A CN111126274A CN 111126274 A CN111126274 A CN 111126274A CN 201911348895 A CN201911348895 A CN 201911348895A CN 111126274 A CN111126274 A CN 111126274A
- Authority
- CN
- China
- Prior art keywords
- network
- target population
- character
- current
- features
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
- G06V20/53—Recognition of crowd images, e.g. recognition of crowd congestion
-
- 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
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The application provides a method, a device, equipment and a medium for detecting an entry target crowd, wherein the method comprises the following steps: establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network; acquiring current character characteristics in a current monitoring image; determining a current target crowd corresponding to the current character characteristic according to the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population. Target crowds and non-target crowds are distinguished more effectively, the difficulty in identifying specific characters or crowds by customs personnel is reduced, and the working efficiency is improved.
Description
Technical Field
The present application relates to the field of entry and exit detection, and more particularly, to a method, an apparatus, a device, and a medium for detecting entry target people.
Background
At present, the port field law enforcement work is easy to have the following contradictions: one is the contradiction between the increasing workload and the lack of law enforcement personnel. In the existing inspection mode, passenger carried objects are taken as marks, all articles of the inbound passengers are objects for inspection and supervision, and the indifference law enforcement makes the contradiction between the increasing workload and the relative shortage of the number of law enforcement personnel increasingly prominent; the second is the contradiction between smooth clearance and strict clearance. Shenzhen customs guarantees certain difficulty and obstacle in strict law enforcement under the big background of facilitated customs; thirdly, the contradiction between more resources and low law enforcement efficiency. At present, the traditional inspection mode lacks pertinence and scientific classification, and the subjective selectivity in the aspects of law enforcement force, law enforcement mode, law enforcement object, law enforcement time and the like is larger.
Meanwhile, the separation of the checking of 'people' and 'things' causes the luck psychology of passengers, and the illegal carrying of the passengers is repeated and checked, and the repeated checking and prohibition are difficult to completely eliminate. In addition, in the field check of the carried articles of the port passengers, the situation of active declaration of the passengers is very few, and port workers mainly check whether the passengers carry strictly prohibited articles or not in the field by means of inquiry, visual observation, packet opening and spot check and the like according to experience. Therefore, the inspection work is blind, and serious missing inspection is easy to cause. From the perspective of risk analysis, because a considerable part of the harmful organisms is not detected, and further the quarantine entry is not passed, the possibility of carrying harmful organisms and epidemic diseases of animals and plants is greatly increased, and the social security is threatened.
At present, customs applies a 'one-machine two-screen' X-ray machine inspection mode on the port site, so that large articles of forbidden articles are effectively prevented from entering the port, but the coping strategies of 'water passengers' (illegally carrying forbidden articles to enter the port) are obviously changed. Firstly, the form of carrying articles is changed, the articles are carried in batches by adopting a small quantity of portable plastic bags, backpacks and the like to avoid X-ray machine screening, or machine inspection and check are carried out artificially and deliberately without an X-ray machine, and no specific effective measure exists at home and abroad at present; and secondly, enlisting a 'new soldier' (illegally carrying prohibited goods inbound personnel without related records for the first time) to greatly increase the discrimination difficulty in the checking process, wherein the randomly drawn customs records the inbound and outbound records of 100 'water customers' carrying a large amount of frozen products of seafood and the like in case of 100, and 42 persons, 51 persons after 1-3 times and only 7 persons more than 3 times which are not intercepted and recorded at the port are found. Therefore, passenger carrying object checking equipment based on the infrared imaging technology is deployed, a good restraining effect is achieved, but the checking equipment is only effective on passenger luggage carrying low-temperature frozen objects, and the problem that 'water customers' gange and use 'new soldiers' to get luck is useless and cannot be radically solved.
Disclosure of Invention
In view of the above, the present application is directed to a method, apparatus, device and medium for detecting an inbound target population that overcomes or at least partially solves the above problems, comprising:
a method for detecting an inbound target population, comprising:
establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network;
acquiring current character characteristics in a current monitoring image;
determining a current target crowd corresponding to the current character characteristic according to the corresponding relation; specifically, determining a current target group corresponding to the character features comprises: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
Further, the air conditioner is provided with a fan,
the character features include: human body features and/or environment features, and/or a one-dimensional or more than two-dimensional array consisting of features extracted from the human body features and the environment features according to a set rule; wherein the content of the first and second substances,
the human body features include: face gray, face texture, face shape, facial expression, head pose, number of following people;
and/or the presence of a gas in the gas,
the environmental characteristics include: illumination intensity, decoration;
and/or the presence of a gas in the gas,
the corresponding relation comprises: a functional relationship; the character features are input parameters of the functional relationship, and the target population is output parameters of the functional relationship;
determining a current target population corresponding to the current character characteristics, further comprising:
and when the corresponding relation comprises a functional relation, inputting the current character characteristics into the functional relation, and determining the output parameters of the functional relation as the current target population.
Further, the step of establishing a corresponding relationship between the character features and the target population includes:
obtaining sample data for establishing a correspondence between the human features and the target population;
analyzing the characteristics and the rules of the character characteristics, and determining the network structure and the network parameters of the artificial neural network according to the characteristics and the rules;
and training and testing the network structure and the network parameters by using the sample data, and determining the corresponding relation between the character characteristics and the target population.
Further, the step of obtaining sample data for establishing a correspondence between the human character features and the target population includes:
collecting the human features and the target population of patients of different lung nodule conditions;
analyzing the character features, and selecting data related to the target population as the character features by combining prestored expert experience information;
and taking the data pair formed by the target population and the selected human character characteristics as sample data.
Further, the air conditioner is provided with a fan,
the Network structure comprises at least one of a CNN Network, a Faster R-CNN Network, an FPN Network, an SqeezeNet Network, a VGG model, a GoogleNet Network, a ResNet Network and a Network-In-Network model;
and/or the presence of a gas in the gas,
the network parameters comprise: at least one of the number of dense blocks, the number of output layers, the number of convolution layers, the number of deconvolution layers, the number of transition layers, the initial weight, and the offset value.
Further, the air conditioner is provided with a fan,
training the network structure and the network parameters, including:
selecting a part of data in the sample data as a training sample, inputting the character features in the training sample into the network structure, and training through an activation function of the network structure and the network parameters to obtain an actual training result;
determining whether an actual training error between the actual training result and a corresponding target population in the training sample meets a preset training error;
determining that the training of the network structure and the network parameters is completed when the actual training error meets the preset training error;
and/or the presence of a gas in the gas,
testing the network structure and the network parameters, comprising:
selecting another part of data in the sample data as a test sample, inputting the character features in the test sample into the trained network structure, and testing by using the activation function and the trained network parameters to obtain an actual test result;
determining whether an actual test error between the actual test result and a corresponding target population in the test sample satisfies a set test error;
and when the actual test error meets the set test error, determining that the test on the network structure and the network parameters is finished.
Further, the air conditioner is provided with a fan,
training the network structure and the network parameters, further comprising:
when the actual training error does not meet the set training error, updating the network parameters through an error energy function of the network structure;
retraining through the activation function of the network structure and the updated network parameters until the retrained actual training error meets the set training error;
and/or the presence of a gas in the gas,
testing the network structure and the network parameters, further comprising:
and when the actual test error does not meet the set test error, retraining the network structure and the network parameters until the retrained actual test error meets the set test error.
A device for detecting an inbound target population, comprising:
the establishing module is used for establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network;
the acquisition module is used for acquiring the current character characteristics in the current monitoring image;
the determining module is used for determining the current target crowd corresponding to the current character characteristic through the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
An apparatus comprising a processor, a memory and a computer program stored on the memory and capable of running on the processor, the computer program when executed by the processor implementing the steps of the method of detecting an inbound target population as described above.
A computer readable storage medium having stored thereon a computer program which, when executed by a processor, carries out the steps of the method of detecting an inbound target population as described above.
The application has the following advantages:
in the embodiment of the application, the corresponding relation between the target population and the character characteristics in the monitored image is established by utilizing the self-learning capability of the artificial neural network; acquiring current character characteristics in a current monitoring image; determining a current target person group corresponding to the current person characteristics according to the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features in the corresponding relationship, which are the same as the current character features, as the current target population, so that the target population and the non-target population are distinguished more effectively, the difficulty in identifying specific characters or populations by customs personnel is reduced, and the working efficiency is improved.
Drawings
In order to more clearly illustrate the technical solutions of the present application, the drawings needed to be used in the description of the present application will be briefly introduced below, and it is apparent that the drawings in the following description are only some embodiments of the present application, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive labor.
FIG. 1 is a flowchart illustrating steps of a method for detecting an inbound target group according to an embodiment of the present disclosure;
fig. 2 is a flowchart illustrating a face detection method for an inbound target group according to an embodiment of the present disclosure;
fig. 3 is a block diagram illustrating a detecting apparatus for an inbound target group according to an embodiment of the present disclosure;
fig. 4 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present application more comprehensible, the present application is described in further detail with reference to the accompanying drawings and the detailed description. It is to be understood that the embodiments described are only a few embodiments of the present application and not all 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 application.
It should be noted that, in any embodiment of the present invention, the artificial neural network is a deep learning target detection network, wherein a backbone network (backbone) in the artificial neural network is used to extract image features, and since a lung CT image contains a very small lesion region, the conventional convolution in the backbone is modified to include convolution of spatial information and channel information. By the method, the network model can pay more attention to the related features, and the attention degree to the unrelated features is reduced. For the detection of pulmonary nodules, deconvolution and transverse connection layers are introduced, and the positions of the nodules are respectively predicted on feature maps with three different resolutions after deconvolution so as to reduce the rate of missed diagnosis of the nodules.
Referring to fig. 1, a method for detecting an entry target group according to an embodiment of the present application is shown, including:
s110, establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network;
s120, obtaining the current character characteristics in the current monitoring image;
s130, determining a current target person group corresponding to the current person characteristic according to the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
In the embodiment of the application, the corresponding relation between the target population and the character characteristics in the monitored image is established by utilizing the self-learning capability of the artificial neural network; acquiring current character characteristics in a current monitoring image; determining a current target person group corresponding to the current person characteristics according to the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features in the corresponding relationship, which are the same as the current character features, as the current target population, so that the target population and the non-target population are distinguished more effectively, the difficulty in identifying specific characters or populations by customs personnel is reduced, and the working efficiency is improved.
The implementation of the crowd inspection by adopting the equipment for realizing the method is favorable for controlling the behavior of carrying harmful substances to invade by 'water passengers' in all the country at the inspection ports of the inbound target crowd, implements more accurate striking on the inspection of the inbound and outbound passengers and the inbound and outbound cargos, stops the occurrence of illegal behaviors such as smuggling and the like, and firmly maintains the national security.
Next, a method of detecting the entry target group in the present exemplary embodiment will be further described.
As described in step S110, the self-learning capability of the artificial neural network is used to establish the corresponding relationship between the target person group and the person feature in the monitored image.
For example: and analyzing the display state rule of the monitoring image corresponding to the target crowd by utilizing an artificial neural network algorithm, and finding out the mapping rule between the patient character characteristic and the target crowd through the self-learning and self-adaptive characteristics of the artificial neural network.
For example: the artificial neural network algorithm can be utilized, the human characteristics of a large number of different volunteers (including but not limited to one or more of age, gender, number of crowd persons, crowd density and the like) are collected in a gathering mode, the human characteristics of a plurality of volunteers and target crowd are selected as sample data, the neural network is learned and trained, the neural network is enabled to fit the relationship between the human characteristics and the target crowd by adjusting the weight between the network structure and the network nodes, and finally the neural network can accurately fit the corresponding relationship between the human characteristics of different patients and the target crowd.
In one embodiment, the character features include: human body features and/or environmental features, and/or a one-dimensional or more than two-dimensional array consisting of features extracted from the human body features and the environmental features according to a set rule;
optionally, the human body features include: face gray, face texture, face shape, facial expression, head pose, number of following people;
optionally, the environmental characteristics include: illumination intensity, decoration;
in an embodiment, the correspondence includes: and (4) functional relation.
Preferably, the character features are input parameters of the functional relationship, and the target population is output parameters of the functional relationship;
therefore, through the corresponding relations in various forms, the flexibility and convenience of determining the current target crowd can be improved.
It should be noted that, because many 'water customers' often adopt team operation nowadays, there are many 'water customers' carrying illegal articles together at the same time to pass customs, and the water customers and the post-screen objects are not all in the existing entry target crowd base at present, and need to carry out analysis and control of the same person according to the early-warning 'water customers', so that customs officers are effectively helped to identify the current 'undiscovered' target crowd.
When a certain 'water visitor' is found through face gray scale, face texture, face shape, face expression and head posture, a face picture of the 'water visitor' is obtained, and big data analysis is carried out after the face picture of the 'water visitor' is modeled to obtain the identity ID of the passenger. Searching a specific place and a specific time point which appear within preset time in a database (historical record) by using the ID, searching all passengers which appear within the preset time according to the time and the place, counting the times of the same passengers, and listing the passengers of the same passers from at least according to the times, wherein the listed passengers of the same passers can be secondarily analyzed and confirmed by customs personnel, and the face picture information is stored and used as a target group identification target. And after that, when the ' water visitor ' appears again, alarming treatment is carried out to remind customs personnel to carry out on-site interception on the ' water visitor ' and the personnel following the water visitor '.
In an embodiment, the specific process of "establishing the correspondence between the character features and the target group" in step S110 may be further described with reference to the following description.
The following steps are described: acquiring sample data for establishing a corresponding relation between the character characteristics and the target population;
in an advanced embodiment, a specific process of acquiring sample data for establishing a corresponding relationship between the human features and the target population may be further described in conjunction with the following description.
The following steps are described: collecting the human features and the target population for patients of different lung nodule conditions;
for example: data collection: collecting character characteristics of patients with different health conditions and corresponding target people groups; and collecting the character characteristics of patients of different ages and corresponding target population; and collecting the character characteristics of the patients with different genders and the corresponding target population.
Therefore, the operation data are collected through multiple ways, the quantity of the operation data is increased, the learning capacity of the artificial neural network is improved, and the accuracy and the reliability of the determined corresponding relation are improved.
The following steps are described: analyzing the character features, and selecting data related to the target population as the character features by combining with prestored expert experience information (for example, selecting character features influencing the target population as input parameters, and using specified parameters as output parameters);
for example: the human characteristics in the relevant data of the diagnosed volunteers are used as input parameters, and the target population in the relevant data is used as output parameters.
The following steps are described: and taking the data pair formed by the target crowd and the selected character characteristics as sample data.
For example: and using part of the obtained input and output parameter pairs as training sample data and using part of the obtained input and output parameter pairs as test sample data.
Therefore, the collected character features are analyzed and processed to obtain sample data, the operation process is simple, and the reliability of the operation result is high.
The following steps are described: analyzing the characteristics and the rules of the character characteristics, and determining the network structure and the network parameters of the artificial neural network according to the characteristics and the rules;
for example: according to the data characteristics and the rules of the data characteristics, which have influences on the pulmonary nodule conditions, such as different ages, illness states, sexes and the like, the basic structure of the network, the number of input and output nodes of the network, the number of hidden layers of the network, the number of hidden nodes, the initial weight of the network and the like can be preliminarily determined.
Preferably, the network structure comprises: at least one of a CNN Network, a Faster R-CNN Network, an FPN Network, a SqeezeNet Network, a VGG model, a GoogleNet Network, a ResNet Network, and a Network-In-Network model.
Preferably, the network parameters include: at least one of the number of dense blocks, the number of output layers, the number of convolution layers, the number of deconvolution layers, the number of transition layers, the initial weight, and the offset value.
Referring to fig. 2, as an example, with the increase of the intelligent demand in the fields of traffic safety, video monitoring, and the like, information fusion plays an important role in face recognition and pedestrian detection technologies. The pedestrian detection has high application value in the industries of vehicle auxiliary driving, intelligent robots and the like, and the detection technology based on multi-model and inter-frame information fusion researched in China can make up the loss of missing detection and false detection existing in a single-frame image detection algorithm.
After the Fast R-CNN network is improved, the Fast R-CNN with higher precision and better performance is produced. Fast R-CNN uses a Region pro-social Networks (RPN) to replace the traditional Selective Search (Selective Search), and has better performance aiming at technologies such as target detection algorithm and multi-sensor fusion. Firstly, fusing complementary detection results of Fast R-CNN and Fast R-CNN models to obtain an accurate detection window, wherein the end-to-end model of the Fast R-CNN does not need to extract a candidate window first, and can directly generate a final window, so that the detection speed can be accelerated, and meanwhile, due to the sharing of the Fast R-CNN network structure, the network can obtain a better detection result; and then, correcting the detection window by adopting a video interframe context fusion algorithm, so that the real-time performance and the accuracy of image detection are improved. Different from the traditional detection algorithm, the fusion technology uses the candidate region algorithm to replace the traditional sliding window strategy, so that the candidate window positioning is more accurate and the time complexity is lower. In addition, the Convolutional Neural Network (CNN) features with stronger robustness are used for replacing the features such as the classical gradient direction Histogram (HOG) and the like, the overall performance of the detection algorithm is improved, and the method is a key technology which is mainly researched and developed at present. The multi-feature fusion algorithm flow is shown in fig. 2. Therefore, when the fast R-CNN network structure is used as the artificial neural network model of the method, the method has the following advantages: when the real-time online comparison is carried out, the human face detection rate is more than 99%, and the correct recognition rate is more than 95%; when the real-time online comparison is carried out, the storage capacity of the face library is large; the adaptability to the posture, angle and expression of the recognition target is good; the correct recognition rate of the photos with large age span (within 20 years) is high; the method has good adaptability to the environment, is stable and reliable, and is suitable for various indoor and outdoor scenes.
It should be noted that the depth of the neural network is not too deep, because the diameter of the lung nodule is generally only 3-30 mm, and the lung nodule only occupies a very small region in the three-dimensional lung CT image, if the number of layers is too large, the information of the lung nodule is greatly weakened, and the detection accuracy is reduced.
Specifically, the features extracted by the conventional convolution are processed in two ways: the first path automatically acquires the importance degree of each feature channel in a learning mode through two operations of compressing and exciting the feature map, promotes useful features according to the importance degree of the feature channels, and inhibits features with small influence on the current task. Wherein the compression operation is achieved using global average pooling; the excitation operation is realized by adopting two layers of full-connection layers, a Relu activation function layer is connected behind the first layer of full-connection layer, a Sigmoid activation function layer is connected behind the second layer of full-connection layer, and the weight is normalized to be 0-1. And the second path compresses the characteristic channels by adopting 1 × 1 convolution and 3 × 3 convolution, and normalizes the weight corresponding to the spatial information to be between 0 and 1 through a Sigmoid activation function layer and fuses.
A convolutional layer containing 3 layers 3 x 3, followed by 6 residual blocks, each of which is then connected to a pooling layer for down-sampling, each residual block consisting of 4 residual units. Since the nodules of 3-10 mm in the data set are numerous, in order to improve the detection accuracy of the network model for the nodules, the feature map after the down-sampling of the fourth layer is up-sampled twice, and the feature map obtained through the up-sampling is transversely connected with the down-sampling feature map with the corresponding size, so that the feature information is fully utilized.
Optionally, a specific process of training the network structure and the network parameters in the step of training and testing the network structure and the network parameters using the sample data to determine the corresponding relationship between the character features and the target population may be further described in conjunction with the following description.
Selecting a part of data in the sample data as a training sample, inputting the character features in the training sample into the network structure, and training through the network structure and the network parameters to obtain an actual training result; determining whether an actual training error between the actual training result and a corresponding target population in the training sample meets a preset training error; determining that the training of the network structure and the network parameters is completed when the actual training error satisfies the preset training error;
more optionally, training the network structure and the network parameters further includes:
when the actual training error does not meet the set training error, updating the network parameters through an error energy function of the network structure; retraining through the activation function of the network structure and the updated network parameters until the retrained actual training error meets the set training error;
for example: and if the test error meets the requirement, finishing the network training test.
Therefore, the reliability of the network structure and the network parameters is further verified by using the test sample for testing the network structure and the network parameters obtained by training.
Optionally, a specific process of testing the network structure and the network parameters in the step of training and testing the network structure and the network parameters using the sample data and determining the corresponding relationship between the character features and the target population may be further described in conjunction with the following description.
Selecting another part of data in the sample data as a test sample, inputting the character features in the test sample into the trained network structure, and testing by using the activation function and the trained network parameters to obtain an actual test result; determining whether an actual test error between the actual test result and a corresponding target population in the test sample satisfies a set test error; and when the actual test error meets the set test error, determining that the test on the network structure and the network parameters is finished.
As described in step S120 above, the current person feature in the current monitored image is obtained;
as described in step S130 above, the current target group corresponding to the current character characteristic is determined through the corresponding relationship.
For example: and identifying the character features in the monitored image in real time.
Therefore, the current target people group is effectively identified according to the current character characteristics based on the corresponding relation, so that accurate judgment basis is provided for diagnosis of doctors, and the judgment result is good in accuracy.
In an alternative example, the step of determining the current target group corresponding to the human character feature in step S130 may include: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
In an optional example, the determining, in step S130, a current target group corresponding to the human character feature may further include: when the corresponding relation can comprise a functional relation, inputting the current character characteristics into the functional relation, and determining the output parameters of the functional relation as the current target population.
Therefore, the current target population is determined according to the current character characteristics based on the corresponding relation or the functional relation, the determination mode is simple and convenient, and the reliability of the determination result is high.
In an alternative embodiment, the method may further include: and verifying whether the current target population is consistent with the actual target population.
Optionally, when a verification result that the current target group does not match the actual target group is received and/or it is determined that there is no person feature in the correspondence that is the same as the current person feature, at least one maintenance operation of updating, correcting, and relearning the correspondence may be performed.
For example: the device itself can not know the actual target crowd, and needs the feedback operation of the doctor, namely, if the device intelligently judges the target crowd, the doctor can know the target crowd by operating the device to feed back that the state of the target crowd is not consistent with the actual state.
And verifying whether the current target crowd is consistent with the actual target crowd (for example, displaying the actual target crowd through an AR display module to verify whether the determined current target crowd is consistent with the actual target crowd).
And when the current target population is not consistent with the actual target population and/or the corresponding relationship does not have the character characteristics which are the same as the current character characteristics, performing at least one maintenance operation of updating, correcting and relearning on the corresponding relationship.
For example: the current target population can be determined according to the maintained corresponding relation and the current character characteristics. For example: and determining the target population corresponding to the character features which are the same as the current character features in the maintained corresponding relationship as the current target population.
Therefore, the corresponding relation between the determined character characteristics and the target crowd is maintained, and the accuracy and the reliability of determining the target crowd are improved.
For the device embodiment, since it is basically similar to the method embodiment, the description is simple, and for the relevant points, refer to the partial description of the method embodiment.
Referring to fig. 3, a device for detecting an entry target group according to an embodiment of the present application is shown, including:
the establishing module 510 is configured to establish a correspondence between the target population and the character features in the monitored image by using the self-learning capability of the artificial neural network;
an obtaining module 520, configured to obtain a current person feature in a current monitored image;
a determining module 530, configured to determine, according to the corresponding relationship, a current target group corresponding to the current character characteristic; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
In one embodiment, the character features include: the motion mode extracted according to a set rule in the interest area image sequence is used for representing each pixel; wherein the content of the first and second substances,
the character features include: human body features and/or environment features, and/or a one-dimensional or more than two-dimensional array consisting of features extracted from the human body features and the environment features according to a set rule; wherein the content of the first and second substances,
the human body features include: face gray, face texture, face shape, facial expression, head pose, number of following people;
and/or the presence of a gas in the gas,
the environmental characteristics include: illumination intensity, decoration;
and/or the presence of a gas in the gas,
the corresponding relation comprises: a functional relationship; the character features are input parameters of the functional relationship, and the target population is output parameters of the functional relationship;
determining a current target population corresponding to the current character characteristics, further comprising:
and when the corresponding relation comprises a functional relation, inputting the current character characteristics into the functional relation, and determining the output parameters of the functional relation as the current target population.
In one embodiment, the establishing module 510 includes:
the acquisition submodule is used for acquiring sample data for establishing the corresponding relation between the character characteristics and the target population;
the analysis submodule is used for analyzing the characteristics and the rules of the character characteristics and determining the network structure and the network parameters of the artificial neural network according to the characteristics and the rules;
and the training submodule is used for training and testing the network structure and the network parameters by using the sample data and determining the corresponding relation between the character characteristics and the target population.
In one embodiment, the obtaining sub-module includes:
a collection sub-module for collecting the human features and the target population for patients of different lung nodule conditions;
the analysis submodule is used for analyzing the character characteristics and selecting data related to the target population as the character characteristics by combining prestored expert experience information;
and the sample data generation submodule is used for taking the target crowd and the data pair formed by the selected character characteristics as sample data.
In one embodiment of the present invention, the substrate is,
the Network structure comprises at least one of a CNN Network, a Fast R-CNN Network, an FPN Network, an SqeezeNet Network, a VGG model, a GoogleNet Network, a ResNet Network and a Network-In-Network model;
and/or the presence of a gas in the gas,
the network parameters comprise: at least one of the number of dense blocks, the number of output layers, the number of convolution layers, the number of deconvolution layers, the number of excess layers, the initial weight, and the offset value.
In one embodiment of the present invention, the substrate is,
the training submodule includes:
a training result generation submodule, configured to select a part of the data in the sample data as a training sample, input the character features in the training sample to the network structure, and perform training through an activation function of the network structure and the network parameters to obtain an actual training result;
the training result error judgment submodule is used for determining whether the actual training error between the actual training result and the corresponding target population in the training sample meets a preset training error or not;
a training completion determination submodule configured to determine that the training of the network structure and the network parameters is completed when the actual training error satisfies the preset training error;
and/or the presence of a gas in the gas,
a test sub-module for testing the network structure and the network parameters, the test sub-module comprising:
a test result generation submodule, configured to select another part of the sample data as a test sample, input the character features in the test sample into the trained network structure, and perform a test with the activation function and the trained network parameters to obtain an actual test result;
the test result error judgment submodule is used for determining whether the actual test error between the actual test result and the corresponding target population in the test sample meets a set test error;
and the test completion judging submodule is used for determining that the test on the network structure and the network parameters is completed when the actual test error meets the set test error.
In one embodiment of the present invention, the substrate is,
the training submodule further comprises:
the network parameter updating submodule is used for updating the network parameters through an error energy function of the network structure when the actual training error does not meet the set training error;
the first retraining submodule is used for retraining through the network structure and the updated network parameters until the retrained actual training error meets the set training error;
and/or the presence of a gas in the gas,
the test submodule further comprises:
and the second retraining submodule is used for retraining the network structure and the network parameters when the actual test error does not meet the set test error until the retrained actual test error meets the set test error.
Referring to fig. 4, a computer device for illustrating a method for detecting an inbound target group according to the present invention may specifically include the following:
the computer device 12 described above is embodied in the form of a general purpose computing device, and the components of the computer device 12 may include, but are not limited to: one or more processors or processing units 16, a system memory 28, and a bus 18 that couples various system components including the system memory 28 and the processing unit 16.
The system memory 28 may include computer system readable media in the form of volatile memory, such as Random Access Memory (RAM)30 and/or cache memory 32. The computer device 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, the storage system 34 may be used to read from and write to non-removable, nonvolatile magnetic media (commonly referred to as "hard drives"). Although not shown in FIG. 4, a magnetic disk drive for reading from and writing to a removable, non-volatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, non-volatile optical disk (e.g., a CD-ROM, DVD-ROM, or other optical media) may be provided. In these cases, each drive may be connected to bus 18 by one or more data media interfaces. The memory may include at least one program product having a set (e.g., at least one) of program modules 42, with the program modules 42 configured to carry out the functions of embodiments of the invention.
Program/utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory, such program modules 42 including, but not limited to, an operating system, one or more application programs, other program modules 42, and program data, each of which examples or some combination thereof may comprise an implementation of a network environment. Program modules 42 generally carry out the functions and/or methodologies of the described embodiments of the invention.
The processing unit 16 executes various functional applications and data processing by running programs stored in the system memory 28, for example, to implement the method for detecting the inbound target group according to the embodiment of the present invention.
That is, the processing unit 16 implements, when executing the program,: establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network; acquiring current character features in a current monitoring image; determining a current target crowd corresponding to the current character characteristic through the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
In an embodiment of the present invention, the present invention further provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor, implements the method for detecting the inbound target group as provided in all embodiments of the present application:
that is, the program when executed by the processor implements: establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network; acquiring current character characteristics in a current monitoring image; determining the current target crowd corresponding to the current character characteristic according to the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
Any combination of one or more computer-readable media may be employed. The computer readable medium may be a computer-readable storage medium or a computer-readable signal medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPOM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider). The embodiments in the present specification are described in a progressive manner, each embodiment focuses on differences from other embodiments, and the same and similar parts among the embodiments are referred to each other.
While preferred embodiments of the present application have been described, additional variations and modifications of these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including the preferred embodiment and all such alterations and modifications as fall within the true scope of the embodiments of the application.
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. 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 terminal 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 terminal. 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 terminal apparatus that comprises the element.
The method, the device, the equipment and the medium for detecting the entry target population provided by the application are introduced in detail, specific examples are applied in the text to explain the principle and the implementation mode of the application, and the description of the above examples is only used for helping to understand the method and the core idea of the application; meanwhile, for the general technical staff in the field, according to the idea of the present application, there may be changes in the specific embodiments and the application scope, and in summary, the content of the present specification should not be understood as the limitation of the present application.
Claims (10)
1. A method for detecting an inbound target population, comprising:
establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network;
acquiring current character characteristics in a current monitoring image;
determining a current target crowd corresponding to the current character characteristic according to the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
2. The method of claim 1,
the character features include: human body features and/or environment features, and/or a one-dimensional or more than two-dimensional array consisting of features extracted from the human body features and the environment features according to a set rule; wherein the content of the first and second substances,
the human body features include: face gray, face texture, face shape, facial expression, head pose, number of following people;
and/or the presence of a gas in the gas,
the environmental characteristics include: illumination intensity, decoration;
and/or the presence of a gas in the gas,
the corresponding relation comprises: a functional relationship; the character features are input parameters of the functional relationship, and the target population is output parameters of the functional relationship;
determining a current target population corresponding to the current character characteristics, further comprising:
and when the corresponding relation comprises a functional relation, inputting the current character characteristics into the functional relation, and determining the output parameter of the functional relation as the current target population.
3. The method of claim 1, wherein the step of establishing a correspondence between the human features and the target population comprises:
acquiring sample data for establishing a corresponding relation between the character characteristics and the target population;
analyzing the characteristics and the rules of the character characteristics, and determining the network structure and the network parameters of the artificial neural network according to the characteristics and the rules;
and training and testing the network structure and the network parameters by using the sample data, and determining the corresponding relation between the character characteristics and the target population.
4. The method of claim 3, wherein the step of obtaining sample data for establishing correspondence between the human character features and the target population comprises:
collecting the human features and the target population of patients of different lung nodule conditions;
analyzing the character features, and selecting data related to the target population as the character features by combining prestored expert experience information;
and taking the target population and the data pair formed by the selected character characteristics as sample data.
5. The method of claim 4,
the Network structure comprises at least one of a CNN Network, a Faster R-CNN Network, an FPN Network, an SqeezeNet Network, a VGG model, a GoogleNet Network, a ResNet Network and a Network-In-Network model;
and/or the presence of a gas in the gas,
the network parameters comprise: at least one of the number of dense blocks, the number of output layers, the number of convolution layers, the number of deconvolution layers, the number of transition layers, the initial weight, and the offset value.
6. The method according to any one of claims 3 to 5,
training the network structure and the network parameters, including:
selecting a part of data in the sample data as a training sample, inputting the character features in the training sample into the network structure, and training through an activation function of the network structure and the network parameters to obtain an actual training result;
determining whether an actual training error between the actual training result and a corresponding target population in the training sample meets a preset training error;
determining that the training of the network structure and the network parameters is completed when the actual training error meets the preset training error;
and/or the presence of a gas in the gas,
testing the network structure and the network parameters, comprising:
selecting another part of data in the sample data as a test sample, inputting the character features in the test sample into the trained network structure, and testing by using the activation function and the trained network parameters to obtain an actual test result;
determining whether an actual test error between the actual test result and a corresponding target population in the test sample satisfies a set test error;
and when the actual test error meets the set test error, determining that the test on the network structure and the network parameters is finished.
7. The method of claim 6,
training the network structure and the network parameters, further comprising:
when the actual training error does not meet the set training error, updating the network parameters through an error energy function of the network structure;
retraining through the activation function of the network structure and the updated network parameters until the retrained actual training error meets the set training error;
and/or the presence of a gas in the gas,
testing the network structure and the network parameters, further comprising:
and when the actual test error does not meet the set test error, retraining the network structure and the network parameters until the retrained actual test error meets the set test error.
8. An apparatus for detecting a target population of inbound destinations, comprising:
the establishing module is used for establishing a corresponding relation between the target population and the character characteristics in the monitored image by utilizing the self-learning capability of the artificial neural network;
the acquisition module is used for acquiring the current character characteristics in the current monitoring image;
the determining module is used for determining the current target crowd corresponding to the current character characteristic through the corresponding relation; specifically, determining a current target population corresponding to the human character features includes: and determining the target population corresponding to the character features which are the same as the current character features in the corresponding relationship as the current target population.
9. An apparatus comprising a processor, a memory, and a computer program stored on the memory and capable of running on the processor, the computer program when executed by the processor implementing the method of any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911348895.7A CN111126274A (en) | 2019-12-24 | 2019-12-24 | Method, device, equipment and medium for detecting inbound target population |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911348895.7A CN111126274A (en) | 2019-12-24 | 2019-12-24 | Method, device, equipment and medium for detecting inbound target population |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111126274A true CN111126274A (en) | 2020-05-08 |
Family
ID=70501781
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911348895.7A Pending CN111126274A (en) | 2019-12-24 | 2019-12-24 | Method, device, equipment and medium for detecting inbound target population |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111126274A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108073883A (en) * | 2016-11-11 | 2018-05-25 | 深圳云天励飞技术有限公司 | Large-scale crowd attribute recognition approach and device |
| CN108986085A (en) * | 2018-06-28 | 2018-12-11 | 深圳视见医疗科技有限公司 | CT image pulmonary nodule detection method, device, equipment and readable storage medium storing program for executing |
| CN109214238A (en) * | 2017-06-30 | 2019-01-15 | 百度在线网络技术（北京）有限公司 | Multi-object tracking method, device, equipment and storage medium |
| CN110084237A (en) * | 2019-05-09 | 2019-08-02 | 北京化工大学 | Detection model construction method, detection method and the device of Lung neoplasm |
| CN110245603A (en) * | 2019-06-12 | 2019-09-17 | 成都信息工程大学 | A kind of group abnormality behavior real-time detection method |
| CN110400298A (en) * | 2019-07-23 | 2019-11-01 | 中山大学 | Detection method, device, equipment and the medium of heart clinical indices |
-
2019
- 2019-12-24 CN CN201911348895.7A patent/CN111126274A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108073883A (en) * | 2016-11-11 | 2018-05-25 | 深圳云天励飞技术有限公司 | Large-scale crowd attribute recognition approach and device |
| CN109214238A (en) * | 2017-06-30 | 2019-01-15 | 百度在线网络技术（北京）有限公司 | Multi-object tracking method, device, equipment and storage medium |
| CN108986085A (en) * | 2018-06-28 | 2018-12-11 | 深圳视见医疗科技有限公司 | CT image pulmonary nodule detection method, device, equipment and readable storage medium storing program for executing |
| CN110084237A (en) * | 2019-05-09 | 2019-08-02 | 北京化工大学 | Detection model construction method, detection method and the device of Lung neoplasm |
| CN110245603A (en) * | 2019-06-12 | 2019-09-17 | 成都信息工程大学 | A kind of group abnormality behavior real-time detection method |
| CN110400298A (en) * | 2019-07-23 | 2019-11-01 | 中山大学 | Detection method, device, equipment and the medium of heart clinical indices |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Ge et al. | Vision-based analysis of small groups in pedestrian crowds | |
| Russakovsky et al. | Detecting avocados to zucchinis: what have we done, and where are we going? | |
| Xiao et al. | Two-dimensional visual tracking in construction scenarios: A comparative study | |
| CN109154976A (en) | Pass through the system and method for machine learning training object classifier | |
| US20210326638A1 (en) | Video panoptic segmentation | |
| Piciarelli et al. | Surveillance-oriented event detection in video streams | |
| JP2015523753A (en) | Track determination of anomalous objects using variational Bayesian expectation maximization based on Gaussian process | |
| CN108171162A (en) | Crowded degree detection method, apparatus and system | |
| CN110147736A (en) | A kind of repetition anomaly detection method based on posture | |
| CN113378675A (en) | Face recognition method for simultaneous detection and feature extraction | |
| CN113052295B (en) | Training method of neural network, object detection method, device and equipment | |
| CN111553321A (en) | Mobile vendor target detection model, detection method and management method thereof | |
| CN105809119A (en) | Sparse low-rank structure based multi-task learning behavior identification method | |
| Hirschorn et al. | Normalizing flows for human pose anomaly detection | |
| CN115294519A (en) | Abnormal event detection and early warning method based on lightweight network | |
| CN115546899A (en) | Examination room abnormal behavior analysis method, system and terminal based on deep learning | |
| Schmuck et al. | GROWL: group detection with link prediction | |
| CN111753587A (en) | Method and device for detecting falling to ground | |
| Zhang | The Understanding of Spatial-Temporal Behaviors | |
| Zhang et al. | A Multiple Instance Learning and Relevance Feedback Framework for Retrieving Abnormal Incidents in Surveillance Videos. | |
| CN111160150A (en) | Video monitoring crowd behavior identification method based on depth residual error neural network convolution | |
| CN115563652A (en) | Track embedding leakage prevention method and system | |
| CN111126274A (en) | Method, device, equipment and medium for detecting inbound target population | |
| CN115527083A (en) | Image annotation method and device and electronic equipment | |
| Patil et al. | Automated Alarm System for Student Anomalous Action Detection in Examination Based on Video Surveillance Using ML Techniques |
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 |