Port cargo real-time monitoring method based on sparse representation

Abstract

The invention discloses a port cargo real-time monitoring intelligent scene adaptation method based on sparse representation, and the method comprises the following steps: arranging a wireless signal transmitter and a receiver outside a cargo box, transmitting a wireless signal with a specific frequency to the cargo box by the signal transmitter, receiving a signal transmitted after the cargo passes through the signal receiver, and transmitting the received signal to the wireless signal receiver; according to the method, quantitative data of shielding effects of different types of goods on radio waves and labels of goods types are obtained, received signal features are analyzed, and matching with a pre-established goods type signal dictionary is carried out by utilizing a sparse representation technology so as to accurately identify the goods types. A dictionary of signal features representing different cargo categories is created, each cargo category corresponds to a specific signal mode in the dictionary, the system compares the received signals with the signal modes in the dictionary, and the specific category of the cargo is determined based on the comparison. According to the method, goods can be efficiently classified, manual box opening inspection in a traditional method is omitted, dependence on manual operation is reduced, high recognition precision and reliability are achieved, and the method is particularly suitable for logistics and transportation industries needing automatic recognition and classification of a large number of goods.

Description

Technical Field

[0001] The invention relates to the technical field of real-time detection of port cargo transportation, and in particular to a real-time monitoring method for port cargo transportation based on sparse representation of video signals. Background Art

[0002] With the development of logistics today, ports, as important nodes in modern logistics systems, are increasing in size and throughput, with a wide variety of goods and a complex and changeable transportation environment. The difficulty and complexity of port cargo management and monitoring have increased accordingly. Traditional monitoring methods rely on manual and mechanical detection, which is time-consuming and labor-intensive when facing large-scale, high-density freight and is easily affected by human factors. Therefore, the use of radio technology to automatically monitor the cargo in the cargo box eliminates the need for traditional manual unpacking inspection, which is beneficial to the accuracy of monitoring and reduces its labor costs.

[0003] At present, common cargo identification technologies in the logistics industry include barcode recognition, radio frequency identification (RFID), and image recognition. Among them, although barcode recognition and RFID technologies have improved the recognition efficiency to a certain extent, they require the cargo to carry corresponding labels. It is relatively complicated to implement them in actual applications. Although image recognition technology can be used for identification through visual information such as monitoring, in a complex transportation environment, factors such as lighting conditions, angle changes, and cargo stacking methods may affect the accuracy of the recognition results.

[0004] Based on this background, the present invention proposes a cargo classification recognition method based on signal sparse representation. By arranging signal transmitters and receivers outside the cargo box, signal information of the passing cargo is obtained, and sparse representation technology is used to classify the cargo in real time. This method can not only improve the recognition speed, but also reduce the dependence on manual operation, and has high practical value and application potential. Summary of the invention

[0005] The purpose of the present invention is to provide a real-time port cargo monitoring method based on sparse representation, which realizes cargo classification through sparse representation analysis of wireless signals, improves the automation level of port logistics, and reduces monitoring costs.

[0006] To achieve the above object, the present invention provides the following solution: The present invention provides a real-time monitoring method for port cargo based on signal sparse representation, comprising the following steps:

[0007] (1) Arrange a signal transmitter and a receiver on the outside of the cargo box. The signal transmitter transmits a wireless signal of a specific frequency into the cargo box, and the signal receiver receives the signal transmitted after passing through the cargo. Obtain quantitative data on the shielding effect of the target on the radio waves when different types of cargo are loaded in the cargo box and label data of the target cargo category, and divide the data to obtain a training sample set and a test sample set; wherein the training sample set includes the received radio data of the target on different cargo categories and the target location label;

[0008] (2) Process the quantitative data in all cases, process the data of each type of goods into a column vector, and construct a dictionary by concatenating all the column vectors in a column-by-column manner to create a dictionary of signal features containing different types of goods. Each type of goods corresponds to a specific signal pattern in the dictionary;

[0009] (3) Construct the target equation, consider the linear fitting error and regularization term of the test signal, and iteratively solve it to obtain the final sparse vector;

[0010] (4) Perform joint analysis and calculation on several maximum values ​​in the sparse vector to obtain the target prediction category.

[0011] The method of obtaining label data and dividing the data in step (1) comprises the following steps:

[0012] 1) According to the actual situation, select the possible object categories, each category corresponds to a unique digital label;

[0013] 2) There are L radio receiving sensors evenly distributed around a monitoring area, with fixed positions, and a radio transmitting sensor with K variable transmitting positions. The wireless communication between a transmitting sensor and a receiving sensor is called a group of links. There are a total of N = K * L groups of links in the entire experimental scenario.

[0014] 3) When the target container arrives at the detection location, the radio transmitting sensor sends a radio signal, and all receiving sensors receive the signal; the transmitting sensor moves to all optional positions and transmits a radio signal, and all receiving nodes receive the arrival signal at the same time; then, the signal strength is extracted from the spectrum of each link as effective information, and each container detection will obtain an information matrix with a dimension size of L rows and K columns; at the same time, the digital label information of the target container when loaded with various goods is saved, which corresponds to the real category information one by one;

[0015] 4) The collected data set is divided into a training sample set and a test sample set in a certain ratio (such as 5:1).

[0016] The column vector concatenation method in step (2) comprises the following steps:

[0017] Take a type of cargo and pack it in a box. Transmit a radio signal for testing. N receivers will receive the signal and convert the obtained information matrix into a column vector with dimension size N. Record its label information as 1. Repeat the measurement τ times and convert them into column vectors to obtain a total of τ column vectors. The label information is Lab1={1,1,…,1}. Repeat the above two steps until all the data of the cargo categories required to construct the dictionary are collected. Concatenate the column vectors at each position in turn to form the final dictionary. Perform bi-norm normalization on each column of the dictionary and concatenate the corresponding label information to form the annotation information.

[0018] The step (3) comprises the following steps:

[0019] Construct the target equation:

[0020]

[0021] The above objective equation contains two parts: fitting error and regularization term. Where y represents a test signal, W is the dictionary matrix, and x is the desired sparse vector. represents the introduced non-convex GMC regularization; α,λ are both positive parameters; z is an auxiliary vector;

[0022] In the process of solving practical problems, since the number of categories of goods in a certain cargo box is always much smaller than the total number of categories of goods, the sparse coding method is used to solve the sparse vector.

[0023] Step (4) comprises the following steps:

[0024] 1) First, find the sparse vector x = {x 1,1 ,x 1,2 ,…,x 1,τ ,x 2,1 ,...x S,τ}, x∈R n ; Sum the x coefficients of each sparse vector Where 1≤p≤S, so we get x * ={x1,...,x p ,...x S}, x * To obtain the weight vector corresponding to different labels, the larger the weight, the greater the probability of predicting the current position, and x * The elements in are arranged in descending order;

[0025] 2) Extract several maximum values ​​in the weight vector, where each weight corresponds to a different product category label;

[0026] 3) Normalize several weights and perform joint calculations to obtain the predicted target category.

[0027] The normalization processing method and target category prediction method described in step 3) are:

[0028] Normalization: First, several maximum value vectors are normalized according to Normalize, where num is the number of maximum values, and you can get the normalized weight;

[0029] Predicted target category:

[0030]

[0031] Among them, class represents the category of the final target, a i represents the normalized weight corresponding to position i, and class(i) represents the true category label corresponding to i.

[0032] The present invention discloses the following technical effects:

[0033] This method can quickly process a large number of goods. The sparse representation technology of wireless signals reduces the reliance on manual operations and improves the efficiency of cargo classification. The sparse coding method is used to represent the cargo category model during modeling. The noise is handled by the minimum fitting error in the target equation. Non-convex regularization is introduced to ensure the sparsity of the sparse vector required, while maintaining the convexity of the target equation for fast solution. The wireless signal is not affected by lighting conditions and can be transmitted stably under various environmental conditions. This method can work effectively in different cargo stacking methods and transportation environments and has good adaptability. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0035] Figure 1 It is a schematic diagram of a real-time monitoring model of port cargo based on sparse representation of the present invention;

[0036] Figure 2 Schematic diagram of a real-time port cargo monitoring method based on sparse representation in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0038] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0039] The present invention provides a real-time monitoring method for port cargo based on sparse representation, such as Figure 2 Specifically including:

[0040] The actual cargo classification problem is modeled. In actual problems, the number of cargoes to be tested in a cargo box is often far less than the total number of possible cargo categories. Therefore, the problem is converted into a sparse coding problem during modeling. First, test signals of all cargo categories are collected through external radio sensors to form a dictionary. In this process, sparse coding can be used to obtain sparse vectors containing classification information for the collected test signals. Finally, the joint analysis method is used to predict the cargo category.

[0041] (1) Arrange a signal transmitter and a receiver on the outside of the cargo box. The signal transmitter transmits a wireless signal of a specific frequency into the cargo box, and the signal receiver receives the signal transmitted after passing through the cargo. Obtain quantitative data on the shielding effect of the target on radio waves when the cargo box is loaded with different categories of goods and label data of the target cargo category, and divide the data to obtain a training sample set and a test sample set; wherein the training sample set includes the received radio data of the target on different cargo categories and the target location label;

[0042] (2) Process the quantitative data in all cases, process the data of each type of goods into a column vector, and construct a dictionary by concatenating all the column vectors in a column-by-column manner to create a dictionary of signal features containing different types of goods. Each type of goods corresponds to a specific signal pattern in the dictionary;

[0043] (3) Construct the target equation, consider the linear fitting error and regularization term of the test signal, and iteratively solve it to obtain the final sparse vector;

[0044] (4) Perform joint analysis and calculation on several maximum values ​​in the sparse vector to obtain the target prediction category.

[0045] The method of obtaining label data and dividing the data in step (1) comprises the following steps:

[0046] 1) According to the actual situation, select the possible object categories, each category corresponds to a unique digital label;

[0047] 2) If Figure 1 ,L radio receiving sensors are evenly distributed around a monitoring area and are fixed in position, and a radio transmitting sensor is configured with K variable transmitting positions. The wireless communication between a transmitting sensor and a receiving sensor is called a set of links. There are a total of N = K * L sets of links in the entire

[0048] 3) When the target container arrives at the detection location, the radio transmitting sensor sends a radio signal, and all receiving sensors receive the signal; the transmitting sensor moves to all optional positions and transmits a radio signal, and all receiving nodes receive the arrival signal at the same time; then, the signal strength is extracted from the spectrum of each link as effective information, and each container detection will obtain an information matrix with a dimension size of L rows and K columns; at the same time, the digital label information of the target container when loaded with various goods is saved, which corresponds to the real category information one by one;

[0049] 4) The collected data set is divided into a training sample set and a test sample set in a ratio of 5:1.

[0050] The column vector concatenation method in step (2) comprises the following steps:

[0051] Take a type of cargo and pack it in a box. Transmit a radio signal for testing. N receivers will receive the signal and convert the obtained information matrix into a column vector with dimension size N. Record its label information as 1. Repeat the measurement τ times and convert them into column vectors to obtain a total of τ column vectors. The label information is Lab1={1,1,…,1}. Repeat the above two steps until all the data of the cargo categories required to construct the dictionary are collected. Concatenate the column vectors at each position in turn to form the final dictionary. Perform bi-norm normalization on each column of the dictionary and concatenate the corresponding label information to form the annotation information.

[0052] The step (3) comprises the following steps:

[0053] Construct the target equation:

[0054] The above objective equation contains two parts: fitting error and regularization term. Where y represents a test signal, W is the dictionary matrix, and x is the desired sparse vector. represents the introduced non-convex GMC regularization; α,λ are both positive parameters; z is an auxiliary vector;

[0056] In the process of solving practical problems, since the number of categories of goods in a certain cargo box is always much smaller than the total number of categories of goods, the sparse coding method is used to solve the sparse vector.

[0057] Step (4) comprises the following steps:

[0058] 1) First, find the sparse vector x = {x 1,1 ,x 1,2 ,…,x 1,τ ,x 2,1 ,...x S,τ}, x∈R n ; Sum the x coefficients of each sparse vector Where 1≤p≤S, so we get x * ={x1,...,x p ,...x S}, x * To obtain the weight vector corresponding to different labels, the larger the weight, the greater the probability of predicting the current position, and x * The elements in are arranged in descending order;

[0059] 2) Extract several maximum values ​​in the weight vector, where each weight corresponds to a different product category label;

[0060] 3) Normalize several weights and perform joint calculations to obtain the predicted target category.

[0061] The normalization processing method and target category prediction method described in step 3) are:

[0062] Normalization: First, several maximum value vectors are normalized according to Normalize, where num is the number of maximum values, and you can get the normalized weight;

[0063] Predicted target category:

[0064]

[0065] Among them, class represents the category of the final target, a i It represents the normalized weight corresponding to position i, and class(i) represents the true category label corresponding to category i.

[0066] Finally, it should be noted that the above-described embodiments are only specific implementations of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The protection scope of the present invention is not limited thereto. Although the present invention is described in detail with reference to the above-described embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-described embodiments within the technical scope disclosed by the present invention, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention. They should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A real-time port cargo monitoring system based on signal sparse representation, characterized by: The following steps are involved: (1) obtaining quantitative data on the shielding effect of cargo boxes loaded with different types of cargo on radio waves and label data of cargo categories, and dividing the data into a training sample set and a test sample set; wherein the training sample set includes radio receiving data of cargo boxes of different cargo categories and cargo category labels; (2) Process the quantitative data of all cargo categories, and process the data of each category into a column vector. All the column vectors obtained above are concatenated into a dictionary in a column-by-column manner; (3) Construct the target equation, consider the linear fitting error and regularization term of the test signal, and iteratively solve it to obtain the final sparse vector; (4) Perform joint analysis and calculation on several maximum values ​​in the sparse vector to obtain the category of the goods to be detected.

2. The real-time port cargo monitoring system based on signal sparse representation according to claim 1 is characterized in that: 1) The goods are classified into categories, and each category corresponds to a unique digital label; 2) L radio receiving sensors are evenly distributed outside the cargo box, with fixed positions, and a radio transmitting sensor is configured, and K types of cargo are selected as experimental objects. There are a total of N = K * L types of possible signal data in the entire experimental scene; 3) The category of the goods is unknown. A radio transmitting sensor is used to send a radio signal, and all receiving sensors receive the signal. Experiments are conducted on multiple categories of goods. Radio signals are transmitted, and all receiving nodes receive the arriving signals at the same time. Then, the signal strength is extracted from the spectrum of each link as effective information. Each type of goods will get an information matrix with a dimension of L rows and K columns. At the same time, the digital label information representing the category of the goods is saved, corresponding to the information it actually represents. 4) Divide the collected data set into a training sample set and a test sample set in a certain ratio (such as 5:1).

3. The real-time port cargo monitoring system based on signal sparse representation according to claim 1 is characterized in that: The column concatenation method of the column vector is: Take a type of cargo and pack it in a box. Transmit a radio signal for testing. N receivers will receive the signal and convert the obtained information matrix into a column vector with dimension size N. Record its label information as 1. Repeat the measurement τ times and convert them into column vectors to obtain a total of τ column vectors. The label information is Lab1={1,1,…,1}. Repeat the above two steps until all the data of the cargo categories required to construct the dictionary are collected. Concatenate the column vectors at each position in turn to form the final dictionary. Perform bi-norm normalization on each column of the dictionary and concatenate the corresponding label information to form the annotation information.

4. The real-time port cargo monitoring system based on signal sparse representation according to claim 1 is characterized in that: The bi-norm normalization method of the column vector x is: in 5. The real-time port cargo monitoring system based on signal sparse representation according to claim 1 is characterized in that: The step (3) comprises the following steps: 1) Construct the target equation: The above objective equation contains two parts: fitting error and regularization term. Where y represents a test signal, W is the dictionary matrix, and x is the desired sparse vector. Represents the introduced non-convex GMC regularization; α,λ are both positive parameters; z is an auxiliary vector; 2) In the process of solving practical problems, since the number of categories of goods in a certain cargo box is always much smaller than the total number of categories of goods, the sparse coding method is used to solve the sparse vector.

6. The real-time port cargo monitoring system based on signal sparse representation according to claim 5 is characterized in that: The sparse coding solution adopts a forward and backward splitting algorithm. The forward splitting improves the sparse representation by gradually adding new basis vectors, and the backward splitting improves the coding quality by optimizing the existing basis vector set. The two steps are performed alternately until the convergence condition is met or the predetermined stopping criterion is reached.

7. The method according to claim 1, characterized in that The step (4) comprises the following steps: 1) First, find the sparse vector x = {x 1,1 ,x 1,2 ,…,x 1,τ ,x 2,1 ,...x S,τ }, x∈R n ; Sum the x coefficients of each sparse vector Where 1≤p≤S, so we get x * ={x1,...,x p ,...x S }, x * To obtain the weight vector corresponding to different labels, the larger the weight, the greater the probability of predicting the current position, and x * The elements in are arranged in descending order; 2) Extract several maximum values ​​in the weight vector, where each weight corresponds to a different product category label; 3) Normalize several weights and perform joint calculations to obtain the predicted target category.

8. The method according to claim 7, characterized in that: The normalization processing method and target category prediction method described in step 3) are: Normalization: First, several maximum value vectors are normalized according to Normalize, where num is the number of maximum values, and you can get the normalized weight; Predicted target category: Among them, class represents the category of the final target, a i It represents the normalized weight corresponding to position i, and class(i) represents the true category label corresponding to category i.

Images