CN109341763B - Transportation data acquisition system and method based on Internet of things - Google Patents

Transportation data acquisition system and method based on Internet of things Download PDF

Info

Publication number
CN109341763B
CN109341763B CN201811177443.2A CN201811177443A CN109341763B CN 109341763 B CN109341763 B CN 109341763B CN 201811177443 A CN201811177443 A CN 201811177443A CN 109341763 B CN109341763 B CN 109341763B
Authority
CN
China
Prior art keywords
layer
information
goods
container
sample
Prior art date
Application number
CN201811177443.2A
Other languages
Chinese (zh)
Other versions
CN109341763A (en
Inventor
吴丰铭
黄恒
Original Assignee
广东长盈科技股份有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 广东长盈科技股份有限公司 filed Critical 广东长盈科技股份有限公司
Priority to CN201811177443.2A priority Critical patent/CN109341763B/en
Publication of CN109341763A publication Critical patent/CN109341763A/en
Application granted granted Critical
Publication of CN109341763B publication Critical patent/CN109341763B/en

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01DMEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE; ARRANGEMENTS FOR MEASURING TWO OR MORE VARIABLES NOT COVERED IN A SINGLE OTHER SUBCLASS; TARIFF METERING APPARATUS; MEASURING OR TESTING NOT OTHERWISE PROVIDED FOR
    • G01D21/00Measuring or testing not otherwise provided for
    • G01D21/02Measuring two or more variables by means not covered by a single other subclass

Abstract

The invention provides a transportation data acquisition method based on the Internet of things, which is characterized in that various sensors are utilized to complete detection of the environment in a container, the provided convolutional neural network model is utilized to identify goods carried in the container, such as animals are monitored, the motion tracking of abnormal goods can be carried out, the obtained data is encrypted and transmitted, and an encryption algorithm is selectively selected according to the obtained security level to prevent information leakage.

Description

Transportation data acquisition system and method based on Internet of things

Technical Field

The invention belongs to the technical field of data acquisition and analysis, and particularly relates to a transportation data acquisition system and a transportation data acquisition method based on the Internet of things of a shipping system.

Background

With the continuous expansion of the offshore silk road in China, the ocean shipping requirements of China and areas such as Europe, Africa, south America and the like are more and more increased. Shipping has the advantage of large capacity, but the shipping cycle is long. If the loaded import and export agricultural products or animals and plants cannot be monitored in real time, the transported living plants can wither or die, and the transported living animals can die because the transported living animals cannot be found and treated in time in the case of diseases. For example: in some cases, 100 kangaroos are transported from Australia to the republic of China or 80 ostriches are transported from Africa to the republic of China, even other animals or weapons and the like which are required for confidentiality cannot be achieved by land transportation or air transportation, and ocean shipping can be regarded as the only option. At this time, the real-time acquisition and monitoring of the information in the container become important, which is the basis for ensuring the safety of the transported goods. Based on the method, a transportation data acquisition method based on the Internet of things is provided.

The invention creatively utilizes the proposed deep learning algorithm to identify the individual information of the goods and the quantity of the same kind of goods, can also track the motion of the identified abnormal animals, and highlights the identified abnormal animals to corresponding responsible personnel, adopts a brand-new convolutional neural network structure and a pooling method, the network structure is not disclosed and used by the prior art, the method for setting each layer in the network structure and updating the model learning is the first creation of the invention, and is applied to the collection of the container goods of ocean shipping for the first time, the real-time identification and the dynamic tracking of each goods are also realized, and the information security in the ocean shipping is ensured by the newly proposed encryption technology.

Disclosure of Invention

The invention aims to provide a transportation data acquisition system and a transportation data acquisition method based on the Internet of things, which are used for acquiring and processing ship information in real time through an information acquisition part, so that the safety management of goods in a ship is realized. The data acquisition system disclosed by the invention can firstly send information such as a temperature sensor, a humidity sensor, a gas concentration sensor and the like to a data processing end in real time based on the Internet of things to carry out environment detection and analysis, so as to complete the basic container environment monitoring problem; the method is characterized in that the method also creatively utilizes the proposed deep learning algorithm to identify the types and the quantity of cargos, even each living animal can also track the motion of the identified abnormal animal and highlight the abnormal animal to the corresponding responsible personnel, the adopted brand-new convolutional neural network structure is not disclosed and used by the prior art, and the method for setting each layer in the network structure and updating model learning is the initiative of the invention and is not used in the collection of container cargos transported by ocean. In addition, in view of information security in marine transportation, especially illegal data acquisition in the face of electronic scouts and spyware in other countries, it is also proposed to encrypt various transmission data in order to secure the security of transported goods.

In order to solve the technical problems, the invention is realized by the following technical scheme:

firstly, the invention provides a transportation data acquisition method based on the Internet of things, which is characterized by comprising the following steps:

the method comprises the steps that a micro camera is used for obtaining visible light images and/or near infrared images of goods images in a shipping container in real time, and individual information of the goods and quantity information of the same goods are obtained through a recognition algorithm, the camera is installed at the top of each container and can rotate 360 degrees around a circular or elliptical orbit to obtain a panoramic video in the container, and the recognition algorithm is used for completing recognition of each goods and statistics of the same goods through a deep learning method;

monitoring temperature state information in the container in real time by using a temperature sensor;

monitoring humidity state information in the container in real time by using a humidity sensor;

acquiring oxygen content information and carbon dioxide concentration information by using a gas concentration sensor;

positioning and displaying position information, running direction information and running speed information of a ship in real time by using a GIS and a satellite positioning system, wherein the satellite positioning system is one or more of Beidou, GPS, Gerners and Galileo;

compressing and storing the information obtained in the steps, sending the information to a local data processing end through an encryption algorithm, carrying out data analysis and control processing by the local data processing end, and remotely sending the processed data to a database server deployed in China;

the encryption algorithm depends on the security level of the transported goods, one of RSA, SHA, MD5 and 3DES algorithms is adopted for general security level, and quantum communication technology is adopted for high-intensity security level to encrypt, so as to prevent illegal information acquisition of spy ships or reconnaissance planes.

Secondly, the invention provides a transportation data acquisition system based on the Internet of things, which is characterized by comprising the following modules:

the image acquisition and identification module is used for acquiring visible light images and/or near infrared images of goods images in the shipping container in real time by using a miniature camera and acquiring individual information of the goods and quantity information of the same kind of goods by using an identification algorithm, the camera is arranged at the top position in each container and can rotate 360 degrees around a circular or elliptical track to acquire a panoramic video in the container, and the identification algorithm is used for completing the identification of each goods and the statistics of the same kind of goods by using a deep learning method;

the temperature sensor acquisition module monitors the temperature state information in the container in real time by using a temperature sensor;

the humidity sensor acquisition module monitors humidity state information in the container in real time by using a humidity sensor;

the gas sensor acquisition module is used for acquiring oxygen content information and carbon dioxide concentration information by using a gas concentration sensor;

the motion state acquisition module is used for positioning and displaying position information, running direction information and running speed information of a ship in real time by utilizing a GIS and a satellite positioning system, wherein the satellite positioning system is one or more of Beidou, GPS, Gerners and Galileo;

compressing and storing the information obtained by the modules, sending the information to a local data processing terminal through an encryption algorithm, carrying out data analysis and control processing by the local data processing terminal, and remotely sending the processed data to a database server deployed in China;

the encryption algorithm depends on the security level of the transported goods, one of RSA, SHA, MD5 and 3DES algorithms is adopted for general security level, and quantum communication technology is adopted for high-intensity security level for encryption to prevent illegal information acquisition of spy ships or reconnaissance planes

Of course, it is not necessary for any product in which the invention is practiced to achieve all of the above-described advantages at the same time.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a schematic diagram of the system of the present invention.

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.

First embodiment

Referring to fig. 1, the present invention is a first embodiment of a transportation data acquisition system based on the internet of things, including the following modules:

the image acquisition and identification module is used for acquiring visible light images and/or near infrared images of goods images in the shipping container in real time by using a miniature camera and acquiring individual information of the goods and quantity information of the same kind of goods by using an identification algorithm, the camera is arranged at the top position in each container and can rotate 360 degrees around a circular or elliptical track to acquire a panoramic video in the container, and the identification algorithm is used for completing the identification of each goods and the statistics of the same kind of goods by using a deep learning method;

the temperature sensor acquisition module monitors the temperature state information in the container in real time by using a temperature sensor;

the humidity sensor acquisition module monitors humidity state information in the container in real time by using a humidity sensor;

the gas sensor acquisition module is used for acquiring oxygen content information and carbon dioxide concentration information by using a gas concentration sensor;

the motion state acquisition module is used for positioning and displaying position information, running direction information and running speed information of a ship in real time by utilizing a GIS and a satellite positioning system, wherein the satellite positioning system is one or more of Beidou, GPS, Gerners and Galileo;

compressing and storing the information obtained by the modules, sending the information to a local data processing terminal through an encryption algorithm, carrying out data analysis and control processing by the local data processing terminal, and remotely sending the processed data to a database server deployed in China;

the encryption algorithm depends on the security level of the transported goods, one of RSA, SHA, MD5 and 3DES algorithms is adopted for general security level, and quantum communication technology is adopted for high-intensity security level to encrypt, so as to prevent illegal information acquisition of spy ships or reconnaissance planes.

Further, the deep learning manner may also use an automatic coding machine, a recurrent neural network, a long-short memory network, and the specific deep neural network model used may also use one or a combination of AlexNet, VGGNet, Google inclusion Net, ResNet, and RCNN, which are already disclosed in the prior art, and for the mature standard network model named by these names, it is not described again here, and all belong to the prior art.

Second embodiment

Based on the second embodiment of the first embodiment, the adopted brand-new convolutional neural network structure and the pooling method are further provided on the basis of the first embodiment, and the structure is not used and is not used in container cargo collection of ocean shipping.

The convolutional neural network structure comprises an input layer, a bidirectional long and short term memory network (BilSTM) layer, a convolutional layer, a pooling layer, a local connecting layer and a full connecting layer, wherein the input layer is used for collecting color information, contour information, texture information and depth information of goods, and the convolutional layer adopts 7 x 7 convolutional kernels and 16 filters; the size of the pooling window of the pooling layer is 3 x 3, and the number of channels is 32; the local connection layer adopts 32 filters, 32 channels and a convolution kernel of 3 x 3; the input of the full connection layer is from the output of the local connection layer; the pooling method of the pooling layer comprises the following steps:

wherein x iseRepresents the output of the current layer, ueRepresenting the input of an activation function, f () representing an activation function, weRepresents the weight of the current layer, phi represents the loss function, xe-1Represents the output of the next layer, beRepresents the offset and δ represents a constant.

Third embodiment

Based on the first embodiment and the second embodiment, a model learning updating method of the convolutional neural network is further provided, and the model updating method is not used and is not used in container cargo collection of ocean shipping. The deep learning method in the recognition module further comprises: learning and updating the convolutional neural network model in the following mode, and mapping the obtained original sample data into 256-dimensional characteristic vectors; calculating a punishment function, wherein the punishment function is formed by fusing a positive feedback function, a first punishment function and a second punishment function, and the positive feedback function is as follows:

where N denotes the size of the sample data set, yiRepresents a sample xiA is the prediction output of the neural network model;

the first penalty function is as follows:

n represents the size of the sample data set, i takes values of 1-N, yiRepresents a sample xiA corresponding label;represents a sample xiAt its label yiB vector of (a) weight(s) includingAnd bjRepresents a sample xiAt its label yiDeviation of (a) from (b)jRepresents the deviation at output node j;

the second penalty function is as follows:

in the formula (I), the compound is shown in the specification,is a sample xiCorresponding to it with tag yiBy the weighted angle of (a) (-)j,iIs a sample xiWith the weight W at the output node jjM is a preset parameter, and m is more than or equal to 1 and less than or equal to 8; k ═ abs (sign (cos θ)j,i))-sign(cosθj,i)(abs(sign(cos2θj,i))-sign(cosθj,i))/2;

The final penalty function is:

wherein lambda is more than or equal to 1 and less than or equal to 8, mu is more than or equal to 0.3 and less than or equal to β and is less than or equal to 0.8.

Fourth embodiment

Based on the second embodiment and the third embodiment, the counting of the number which can be completed through the detection of the contour of the goods in the container is further provided; the goods to be carried can be living animals and plants or videos, clothes, electronic devices, daily commodities, cosmetics and the like, and the goods to be carried can also be ancient biogenetic fossils, salvaged sunken vessel antiques, marine organisms or national confidential species or weapons and the like;

the quantum communication mode is that the data terminal of the container is connected with the quantum server by two channels, one is a quantum channel and is used for transmitting a quantum key, and the other is an internet channel and is used for transmitting encrypted data; a quantum channel.

Fifth embodiment

Based on the second embodiment and the third embodiment, it is further proposed that the abnormal animal can be dynamically tracked by means of motion tracking, and the state of the abnormal animal in the whole container can be located by using rectangular frame tracking. Because the loading of ship is great, the quantity of container is very much, and the goods quantity that every container loaded is very big again, for example, 50 badgers have been loaded to a container, and 100 containers then can hold how big 5000 badgers, can load ten thousand badgers even. At this time, the ordinary human eye is not sufficient to distinguish the state of each abnormal badger. At this moment, can discern unusual badger earlier, then carry out the sign tracking through the form of rectangle frame to it. Common tracking methods include a feature-based tracking method, a model-based tracking method, an area-based tracking method, and a contour-based tracking method. They are not listed here and belong to the prior art.

Sixth embodiment

A transportation data acquisition method based on the Internet of things is provided, which is characterized in that:

the method comprises the steps that a micro camera is used for obtaining visible light images and/or near infrared images of goods images in a shipping container in real time, and individual information of the goods and quantity information of the same goods are obtained through a recognition algorithm, the camera is installed at the top of each container and can rotate 360 degrees around a circular or elliptical orbit to obtain a panoramic video in the container, and the recognition algorithm is used for completing recognition of each goods and statistics of the same goods through a deep learning method;

monitoring temperature state information in the container in real time by using a temperature sensor;

monitoring humidity state information in the container in real time by using a humidity sensor;

acquiring oxygen content information and carbon dioxide concentration information by using a gas concentration sensor;

positioning and displaying position information, running direction information and running speed information of a ship in real time by using a GIS and a satellite positioning system, wherein the satellite positioning system is one or more of Beidou, GPS, Gerners and Galileo;

compressing and storing the information obtained in the steps, sending the information to a local data processing end through an encryption algorithm, carrying out data analysis and control processing by the local data processing end, and remotely sending the processed data to a database server deployed in China;

the encryption algorithm depends on the security level of the transported goods, one of RSA, SHA, MD5 and 3DES algorithms is adopted for general security level, and quantum communication technology is adopted for high-intensity security level to encrypt, so as to prevent illegal information acquisition of spy ships or reconnaissance planes.

Seventh embodiment

On the basis of the sixth embodiment, a novel convolutional neural network structure and a pooling method are further provided, and the structure is not used and is not used in container cargo collection of ocean shipping.

The convolutional neural network structure comprises an input layer, a bidirectional long and short term memory network (BilSTM) layer, a convolutional layer, a pooling layer, a local connecting layer and a full connecting layer, wherein the input layer is used for collecting color information, contour information, texture information and depth information of goods, and the convolutional layer adopts 7 x 7 convolutional kernels and 16 filters; the size of the pooling window of the pooling layer is 3 x 3, and the number of channels is 32; the local connection layer adopts 32 filters, 32 channels and a convolution kernel of 3 x 3; the input of the full connection layer is from the output of the local connection layer; the pooling method of the pooling layer comprises the following steps:

wherein x iseRepresents the output of the current layer, ueRepresenting the input of an activation function, f () representing an activation function, weRepresents the weight of the current layer, phi represents the loss function, xe-1Represents the output of the next layer, beRepresents the offset and δ represents a constant.

Eighth embodiment

Based on the sixth embodiment and the seventh embodiment, a model learning and updating method of the convolutional neural network is further provided, and the model updating method is not used and is not used in container cargo collection of ocean shipping. The deep learning method in the recognition module further comprises: learning and updating the convolutional neural network model in the following mode, and mapping the obtained original sample data into 256-dimensional characteristic vectors; calculating a punishment function, wherein the punishment function is formed by fusing a positive feedback function, a first punishment function and a second punishment function, and the positive feedback function is as follows:

where N denotes the size of the sample data set, yiRepresents a sample xiA is the prediction output of the neural network model;

the first penalty function is as follows:

n represents the size of the sample data set, i takes values of 1-N, yiRepresents a sample xiA corresponding label;represents a sample xiAt its label yiB vector of (a) weight(s) includingAnd bjRepresents a sample xiAt its label yiDeviation of (a) from (b)jRepresents the deviation at output node j;

the second penalty function is as follows:

in the formula (I), the compound is shown in the specification,is a sample xiCorresponding to it with tag yiBy the weighted angle of (a) (-)j,iIs a sample xiWith the weight W at the output node jjM is a preset parameter, and m is more than or equal to 1 and less than or equal to 8; k ═ abs (sign (cos θ)j,i))-sign(cosθj,i)(abs(sign(cos2θj,i))-sign(cosθj,i))/2;

The final penalty function is:

wherein lambda is more than or equal to 1 and less than or equal to 8, mu is more than or equal to 0.3 and less than or equal to β and is less than or equal to 0.8.

Ninth embodiment

Based on the seventh embodiment and the eighth embodiment, it is further proposed that the counting of the number can be completed by the detection of the contour of the cargo in the container; the goods to be carried can be living animals and plants or videos, clothes, electronic devices, daily commodities, cosmetics and the like, and can also be ancient biogenetic fossils, salvaged sunken vessel antiques, marine organisms or national confidential species or weapons and the like;

the quantum communication mode is that the data terminal of the container is connected with the quantum server by two channels, one is a quantum channel and is used

In the transmission of the quantum key, the other channel is an internet channel and is used for transmitting the encrypted data; a quantum channel.

Tenth embodiment

Based on the seventh embodiment and the eighth embodiment, it is further proposed that the abnormal animal can be dynamically tracked by means of motion tracking, and the state of the abnormal animal in the whole container can be located by using rectangular frame tracking. Because the loading of ship is great, the quantity of container is very much, and the goods quantity that every container loaded is very big again, for example, 50 badgers have been loaded to a container, and 100 containers then can hold how big 5000 badgers, can load ten thousand badgers even. At this time, the ordinary human eye is not sufficient to distinguish the state of each abnormal badger. At this moment, can discern unusual badger earlier, then carry out the sign tracking through the form of rectangle frame to it. Common tracking methods include a feature-based tracking method, a model-based tracking method, an area-based tracking method, and a contour-based tracking method. They are not listed here and belong to the prior art.

In the description herein, references to the description of "one embodiment," "an example," "a specific example" or the like are intended to mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the invention. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example.

Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing. The computer readable medium may be a computer readable signal medium or a computer readable storage medium. The computer readable storage medium may be, for example, 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 (EPROM 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. 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 integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

A storage medium containing computer executable instructions of the transportation data acquisition method based on the internet of things according to the embodiments, wherein the storage medium stores program instructions capable of implementing the method. The integrated unit implemented in the form of a software functional unit may be stored in a computer readable storage medium. The software functional unit is stored in a storage medium and includes several instructions to enable a computer device (which may be a personal computer, a server, or a network device) or a processor (processor) to execute some steps of the methods according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.

The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, or direct or indirect applications in other related fields, which are made by using the contents of the present specification and the accompanying drawings, are included in the scope of the present invention. The preferred embodiments of the invention disclosed above are intended to be illustrative only. The preferred embodiments are not intended to be exhaustive or to limit the invention to the precise embodiments disclosed. Obviously, many modifications and variations are possible in light of the above teaching. The embodiments were chosen and described in order to best explain the principles of the invention and the practical application, to thereby enable others skilled in the art to best utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims (8)

1. A transportation data acquisition method based on the Internet of things is characterized in that:
the method comprises the steps that a miniature camera is used for obtaining visible light images and/or near infrared images of goods images in shipping containers in real time, and individual information of the goods and quantity information of the same goods are obtained through a recognition algorithm, the camera is installed at the top of each container and can rotate 360 degrees around a circular or elliptical track to obtain panoramic videos in the containers;
the identification algorithm completes the identification of each cargo and the quantity statistics of the same cargo through a deep learning method;
monitoring temperature state information in the container in real time by using a temperature sensor;
monitoring humidity state information in the container in real time by using a humidity sensor;
acquiring oxygen content information and carbon dioxide concentration information by using a gas concentration sensor;
positioning and displaying position information, running direction information and running speed information of a ship in real time by using a GIS and a satellite positioning system, wherein the satellite positioning system is one or more of Beidou, GPS, Gerners and Galileo;
compressing and storing the information obtained in the steps, sending the information to a local data processing end through an encryption algorithm, carrying out data analysis and control processing by the local data processing end, and remotely sending the processed data to a database server deployed in China;
the encryption algorithm depends on the security level of the transported goods, one of RSA, SHA, MD5 and 3DES algorithms is adopted for the common security level, and quantum communication technology is adopted for the high-strength security level for encryption to prevent illegal information acquisition of spy ships or reconnaissance planes;
the deep learning method further comprises: learning and updating the convolutional neural network model in the following way, and mapping the obtained original sample data into 256-dimensional feature vectors; calculating a punishment function, wherein the punishment function is formed by fusing a positive feedback function, a first punishment function and a second punishment function, and the positive feedback function is as follows:
where N denotes the size of the sample data set, yiRepresents a sample xiCorresponding labels, a being of neural network modelPredicting and outputting;
the first penalty function is as follows:
n represents the size of the sample data set, i takes values of 1-N, yiRepresents a sample xiA corresponding label; wyiRepresents a sample xiAt its label yiA b vector comprising byiAnd bj,byiRepresents a sample xiAt its label yiDeviation of (a) from (b)jRepresents the deviation at output node j;
the second penalty function is as follows:
where phi (theta)yi,i)=(-1)kcos(mθyi,i)-2k,θyi,iIs a sample xiCorresponding to it with tag yiBy the weighted angle of (a) (-)j,iIs a sample xiWith the weight W at the output node jjM is a preset parameter, and m is more than or equal to 1 and less than or equal to 8; k ═ abs (sign (cos θ)j,i))-sign(cosθj,i)(abs(sign(cos2θj,i))-sign(cosθj,i))/2;
The final penalty function is:
wherein lambda is more than or equal to 1 and less than or equal to 8, mu is more than or equal to 0.3, and β is more than or equal to 0.8;
the convolutional neural network model comprises an input layer, a bidirectional long-short term memory network (BilSTM) layer, a convolutional layer, a pooling layer, a local connection layer and a full connection layer which are connected in sequence.
2. The transportation data acquisition method based on the internet of things as claimed in claim 1, wherein the deep learning method specifically comprises the following steps: a convolutional neural network structure is adopted, the convolutional neural network structure comprises an input layer, a bidirectional long-short term memory network (BilSTM) layer, a convolutional layer, a pooling layer, a local connecting layer and a full connecting layer which are sequentially connected, the input layer collects color information, contour information, texture information and depth information of goods, and the convolutional layer adopts 7 × 7 convolutional kernels and 16 filters; the size of the pooling window of the pooling layer is 3 x 3, and the number of channels is 32; the local connection layer adopts 32 filters, 32 channels and a convolution kernel of 3 x 3; the input of the full connection layer is from the output of the local connection layer; the pooling method of the pooling layer comprises the following steps:
wherein x iseRepresents the output of the current layer, ueRepresents the input of an activation function, f () represents the activation function, phi represents a loss function, weWeight, x, representing the current layere-1Represents the output of the next layer, beRepresents the offset and δ represents a constant.
3. The method for collecting transportation data based on the internet of things as claimed in claim 1 or 2, wherein the identification algorithm can also complete the counting of the quantity through the detection of the outline of the goods in the container;
the goods are one or more of living animals, plants or food, electronic devices and daily commodities;
the quantum communication mode is that the data terminal of the container is connected with the quantum server by two channels, one is a quantum channel and is used for transmitting a quantum key, and the other is an internet channel and is used for transmitting encrypted data; a quantum channel.
4. The method for collecting transportation data based on the internet of things as claimed in claim 1 or 2, wherein the collection method can also realize motion tracking of specified goods, when the loaded goods are animals, the monitored abnormal animals are dynamically tracked, and the state of the abnormal animals in the whole container is tracked and positioned by using a rectangular frame.
5. The transportation data acquisition system based on the Internet of things is characterized by comprising the following modules:
the image acquisition and identification module is used for acquiring visible light images and/or near infrared images of goods images in the shipping container in real time by using a miniature camera and acquiring individual information of the goods and quantity information of the same kind of goods by using an identification algorithm, the camera is arranged at the top position in each container and can rotate 360 degrees around a circular or elliptical track to acquire a panoramic video in the container, and the identification algorithm is used for completing the identification of each goods and the statistics of the same kind of goods by using a deep learning method;
the temperature sensor acquisition module monitors the temperature state information in the container in real time by using a temperature sensor;
the humidity sensor acquisition module monitors humidity state information in the container in real time by using a humidity sensor;
the gas sensor acquisition module is used for acquiring oxygen content information and carbon dioxide concentration information by using a gas concentration sensor;
the motion state acquisition module is used for positioning and displaying position information, running direction information and running speed information of a ship in real time by utilizing a GIS and a satellite positioning system, wherein the satellite positioning system is one or more of Beidou, GPS, Gerners and Galileo;
compressing and storing the information obtained by the modules, sending the information to a local data processing terminal through an encryption algorithm, carrying out data analysis and control processing by the local data processing terminal, and remotely sending the processed data to a database server deployed in China;
the encryption algorithm depends on the security level of the transported goods, one of RSA, SHA, MD5 and 3DES algorithms is adopted for the common security level, and quantum communication technology is adopted for the high-strength security level for encryption to prevent illegal information acquisition of spy ships or reconnaissance planes;
the deep learning method further comprises: learning and updating the convolutional neural network model in the following way, and mapping the obtained original sample data into 256-dimensional feature vectors; calculating a punishment function, wherein the punishment function is formed by fusing a positive feedback function, a first punishment function and a second punishment function, and the positive feedback function is as follows:
where N denotes the size of the sample data set, yiRepresents a sample xiA is the prediction output of the neural network model;
the first penalty function is as follows:
n represents the size of the sample data set, i takes values of 1-N, yiRepresents a sample xiA corresponding label; wyiRepresents a sample xiAt its label yiA b vector comprising byiAnd bj,byiRepresents a sample xiAt its label yiDeviation of (a) from (b)jRepresents the deviation at output node j;
the second penalty function is as follows:
where phi (theta)yi,i)=(-1)kcos(mθyi,i)-2k,θyi,iIs a sample xiCorresponding to it with tag yiBy the weighted angle of (a) (-)j,iIs a sample xiWith the weight W at the output node jjM is a preset parameter, and m is more than or equal to 1 and less than or equal to 8; k ═ abs (sign (cos θ)j,i))-sign(cosθj,i)(abs(sign(cos2θj,i))-sign(cosθj,i))/2;
The final penalty function is:
wherein lambda is more than or equal to 1 and less than or equal to 8, mu is more than or equal to 0.3, and β is more than or equal to 0.8;
the convolutional neural network model comprises an input layer, a bidirectional long-short term memory network (BilSTM) layer, a convolutional layer, a pooling layer, a local connection layer and a full connection layer which are connected in sequence.
6. The internet of things-based transportation data acquisition system according to claim 5, wherein the deep learning method in the identification module specifically comprises: the method comprises the following steps of adopting a convolutional neural network structure, wherein the convolutional neural network structure comprises an input layer, a bidirectional long-short term memory network (BilSTM) layer, a convolutional layer, a pooling layer, a local connecting layer and a full connecting layer, the input layer is used for collecting color information, outline information, texture information and depth information of goods, and the convolutional layer adopts 7 x 7 convolutional kernels and 16 filters; the size of the pooling window of the pooling layer is 3 x 3, and the number of channels is 32; the local connection layer adopts 32 filters, 32 channels and a convolution kernel of 3 x 3; the input of the full connection layer is from the output of the local connection layer; the pooling method of the pooling layer comprises the following steps:wherein x iseRepresents the output of the current layer, ueRepresenting the input of an activation function, f () representing an activation function, weRepresents the weight of the current layer, phi represents the loss function, xe-1Represents the output of the next layer, beRepresents the offset and δ represents a constant.
7. The Internet of things-based transportation data acquisition system as claimed in claim 5 or 6, wherein the identification algorithm can also complete counting of the number through detection of the contour of the goods in the container;
the goods are one or more of living animals, plants or food, electronic devices and daily commodities;
the quantum communication mode is that the data terminal of the container is connected with the quantum server by two channels, one is a quantum channel and is used for transmitting a quantum key, and the other is an internet channel and is used for transmitting encrypted data; a quantum channel.
8. The transportation data acquisition system based on the internet of things as claimed in claim 5 or 6, wherein the acquisition system can also realize motion tracking of specified goods, when the carried goods are animals, the monitored abnormal animals are dynamically tracked, and the state of the abnormal animals in the whole container is tracked and positioned by using a rectangular frame.
CN201811177443.2A 2018-10-10 2018-10-10 Transportation data acquisition system and method based on Internet of things CN109341763B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811177443.2A CN109341763B (en) 2018-10-10 2018-10-10 Transportation data acquisition system and method based on Internet of things

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811177443.2A CN109341763B (en) 2018-10-10 2018-10-10 Transportation data acquisition system and method based on Internet of things

Publications (2)

Publication Number Publication Date
CN109341763A CN109341763A (en) 2019-02-15
CN109341763B true CN109341763B (en) 2020-02-04

Family

ID=65309356

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811177443.2A CN109341763B (en) 2018-10-10 2018-10-10 Transportation data acquisition system and method based on Internet of things

Country Status (1)

Country Link
CN (1) CN109341763B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109842210B (en) * 2019-02-22 2020-02-14 广东科源电气股份有限公司 Monitoring system and method for transformer
CN110243410A (en) * 2019-06-24 2019-09-17 安徽理工大学 A kind of logistics container examination system based on Internet of Things
CN111179133B (en) * 2019-12-30 2020-09-25 智慧校园(广东)教育科技有限公司 Wisdom classroom interaction system

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102830660A (en) * 2011-06-16 2012-12-19 宁波图腾物联科技有限公司 Dangerous goods transportation security detecting method and system thereof
CN104081158B (en) * 2011-11-18 2018-02-02 通腾北美有限公司 For the method and apparatus for the cost data for being created for using in the route across electronic map is produced
CN104200224A (en) * 2014-08-28 2014-12-10 西北工业大学 Valueless image removing method based on deep convolutional neural networks
CN105447458B (en) * 2015-11-17 2018-02-27 深圳市商汤科技有限公司 A kind of large-scale crowd video analytic system and method
CN107358388A (en) * 2016-11-03 2017-11-17 厦门嵘拓物联科技有限公司 A kind of WMS based on Internet of Things and the storage quality risk appraisal procedure based on the system
CN206460325U (en) * 2017-01-23 2017-09-01 成都雅骏新能源汽车科技股份有限公司 A kind of pure electronic logistic car multifunction monitoring system
CN106989691A (en) * 2017-04-01 2017-07-28 成都雅骏新能源汽车科技股份有限公司 A kind of electronic logistic car container intelligent integrated machine
CN107464078A (en) * 2017-08-02 2017-12-12 新疆维吾尔自治区葡萄瓜果研究所 A kind of "Hami" melon Cold Chain Logistics control system and control method
CN207560210U (en) * 2017-11-14 2018-06-29 吉林农业科技学院 Home farm's Environmental security prior-warning device
CN108497839A (en) * 2017-12-18 2018-09-07 上海云拿智能科技有限公司 The shelf of kinds of goods can be perceived
CN107992865A (en) * 2018-01-26 2018-05-04 重庆邮电大学 A kind of vehicle identification method and system based on video analysis

Also Published As

Publication number Publication date
CN109341763A (en) 2019-02-15

Similar Documents

Publication Publication Date Title
Liu et al. Application of deep convolutional neural networks for detecting extreme weather in climate datasets
Hansen et al. Humid tropical forest disturbance alerts using Landsat data
Weinstein A computer vision for animal ecology
O’Mahony et al. Deep learning vs. traditional computer vision
US9488492B2 (en) Real-time system for multi-modal 3D geospatial mapping, object recognition, scene annotation and analytics
Wang et al. Gated convolutional neural network for semantic segmentation in high-resolution images
Talavera et al. Review of IoT applications in agro-industrial and environmental fields
Isaya Ndossi et al. Application of open source coding technologies in the production of land surface temperature (LST) maps from Landsat: a PyQGIS plugin
US10319107B2 (en) Remote determination of quantity stored in containers in geographical region
US10643123B2 (en) Systems and methods for recognizing objects in radar imagery
Lek et al. Artificial neuronal networks: application to ecology and evolution
Johnson et al. Change vector analysis: A technique for the multispectral monitoring of land cover and condition
Corchado et al. Quantifying the Ocean’s CO2 budget with a CoHeL-IBR system
Plaza et al. On the use of small training sets for neural network-based characterization of mixed pixels in remotely sensed hyperspectral images
US20180224854A1 (en) Unmanned vehicle (uv) movement and data control system
Baldi et al. Long-term satellite NDVI data sets: evaluating their ability to detect ecosystem functional changes in South America
Manandhar et al. Improving the accuracy of land use and land cover classification of Landsat data using post-classification enhancement
US7391907B1 (en) Spurious object detection in a video surveillance system
Song Optical remote sensing of forest leaf area index and biomass
Vicharnakorn et al. Carbon stock assessment using remote sensing and forest inventory data in Savannakhet, Lao PDR
Ko et al. Classification of potential water bodies using Landsat 8 OLI and a combination of two boosted random forest classifiers
Hung et al. Multi-class predictive template for tree crown detection
Fensholt et al. Assessing land degradation/recovery in the African Sahel from long-term earth observation based primary productivity and precipitation relationships
McDermid et al. Remote sensing for large-area habitat mapping
Guo et al. Geospatial object detection in high resolution satellite images based on multi-scale convolutional neural network

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
GR01 Patent grant
GR01 Patent grant