# US 2023/0267779 A1 · Method and system for collecting and monitoring vehicle status information

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2023/0267779 A1
- Document ID: US20230267779A1
- Status: Published application
- Kind: A1
- Application number: 17/772,144
- Filed: 2021-06-30
- Published: 2023-08-24
- Earliest claimed: 2021-06-15
- Assignees: Hefei Weitianyuntong Information Technology Co., Ltd
- Inventors: Qing WANG
- Web page: https://patentgrep.com/patent/US20230267779A1

## CPC

- G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69)
- G07C 5/0816: {Indicating performance data, e.g. occurrence of a malfunction}
- G06V 20/625: {License plates}
- G07C 5/0808: {Diagnosing performance data (testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005)}
- G07C 5/12: in graphical form
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G07C: TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE

## Abstract

The present application discloses a method and system for collecting and monitoring vehicle status information. By acquiring vehicle state data and storing it, processing the vehicle state data to obtain optimization data and display it, setting alarm rule, alarm threshold, alarm contact and alarm mode according to the displayed optimization data, and monitoring the optimization data, and generating an alarm when the optimization data reaches the alarm threshold. The vehicle status information collection and monitoring method disclosed in the present application can perform three-dimensional monitoring of resources and display and alarm.

## Claims

### Claim 1

A method for collecting and monitoring vehicle status information, comprising:
acquiring vehicle state data and storing the vehicle state data;
processing the vehicle state data to obtain optimization data and display the optimization data;
setting alarm rule, alarm threshold, alarm contact and alarm mode according to the displayed optimization data; and
monitoring the optimization data, and generating an alarm when the optimization data reaches the alarm threshold.

### Claim 2 (depends on 1)

The method according to claim 1, wherein, the processing of the vehicle state data to obtain the optimization data and display the vehicle state data comprises:
labeling the vehicle state data to obtain thread data;
extracting relevant features of the thread data;
constructing training set and test set of the data according to the relevant features;
constructing a screening model according to the Naive Bayesian classification algorithm, and using the training set and test set to train and test the screening model; and
when the screening model passes training and testing, the optimized data is obtained according to the screening model and is displayed.

### Claim 3 (depends on 2)

The method according to claim 2, wherein, labeling the vehicle state data to obtain thread data comprises:
labeling and classifying the vehicle state data according to the vehicle information, data source, data content and transportation business scenarios that generate data in the vehicle status data; the label comprises valuable data and invaluable data; among them, the valuable data is lead data;
the relevant features of the thread data comprise basic features and associated features; the basic features comprise basic information of the data, and the associated features comprise information related to data acquisition methods.

### Claim 4 (depends on 1)

The method according to claim 1, wherein, the acquiring vehicle state data comprise acquiring vehicle information, comprises:
obtaining vehicle image and extracting license plate image;
transforming the license plate image into a positive-view license plate image; and
using license plate recognition network to recognize positive-view license plate image to obtain vehicle information.

### Claim 5 (depends on 4)

The method according to claim 4, wherein, obtaining the vehicle image and extracting the license plate image comprises:
obtaining the vehicle image, using the YOLOv5 vehicle detection algorithm to detect target for the vehicle in the image, outputting the coordinates of the four corners of the vehicle on the image, and cropping the vehicle image according to the coordinates.

### Claim 6 (depends on 4)

The method according to claim 4, wherein, using license plate recognition network to recognize positive-view license plate image comprises:
using the license plate recognition (LPRnet) wide convolutional neural network to extract features from the positive-view license plate image, to obtain contextual information of local characters; and
decoding the probability of the local characters, finding multiple most probable sequences through beam search, and returning the first sequence that best matches the predefined template set, which is the license plate recognition result.

### Claim 7

A vehicle state information collection and monitoring system, comprising:
data acquisition module, used to acquire vehicle state data, normalize and store;
data processing module, used to process the vehicle state data, obtain optimization data and display the optimization data;
alarm setting module, used to set alarm rule, alarm threshold, alarm contact and alarm method according to the displayed optimization data; and
monitoring module, used to monitor the optimization data and generate an alarm when the optimization data reaches the alarm threshold.

### Claim 8 (depends on 7)

The system according to claim 7, wherein, the data processing module comprises:
screening unit, used to construct a data screening model to screen the vehicle state data;
labeling unit, used to label the vehicle state data to obtain thread data;
extraction unit, used to extract relevant feature of the thread data;
training and testing unit, used to construct the training set and test set of the data according to the relevant feature, and train and test the screening model; and
when the screening model passes training and testing, the optimization is obtained according to the screening model and displayed.

### Claim 9 (depends on 8)

The system according to claim 8, wherein, the labeling unit performs labeling processing on the vehicle state data, and obtaining the thread data, comprises:
labeling and classifying the vehicle state data according to the vehicle information, data source, data content and transportation business scenarios that generate data in the vehicle status data; the label comprises valuable data and invaluable data; among them, the valuable data is lead data;
the relevant features of the thread data comprise basic features and associated features; the basic features comprise basic information of the data, and the associated features comprise information related to data acquisition methods.

### Claim 10 (depends on 7)

The system according to claim 7, wherein, the acquiring vehicle state data comprises acquiring vehicle information, the data acquisition module comprises:
image acquisition unit, used to obtain the vehicle image, and use the YOLOv5 vehicle detection algorithm to extract the license plate image;
image transformation unit, used to transform the license plate image into a positive-view license plate image; and
image recognition unit, used to recognize the positive-view license plate image by using the license plate recognition network,
wherein the image recognition unit recognizes the positive-view license plate image by using the license plate recognition network, comprises:
using the license plate recognition (LPRnet) wide convolutional neural network to extract features from the positive-view license plate image, to obtain contextual information of local characters; and
decoding the probability of the local characters, finding multiple most probable sequences through beam search, and returning the first sequence that best matches the predefined template set, which is the license plate recognition result.