VEHICLE MODEL IDENTIFICATION DEVICE, VEHICLE MODEL IDENTIFICATION SYSTEM, AND VEHICLE MODEL IDENTIFICATION METHOD

• US 20190114494A1
• Filed: 02/15/2017
• Published: 04/18/2019
• Est. Priority Date: 03/29/2016
• Status: Abandoned Application

• Current Assignee: Panasonic Intellectual Property Management Co., Ltd. (Panasonic Holdings Corporation)
• Original Assignee: Panasonic Intellectual Property Management Co., Ltd. (Panasonic Holdings Corporation)
• Inventors: UETA, Daisuke
• Application Number:
  US16/088,546
• Publication Number:
  US 20190114494A1
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/532 Query formulation, e.g. graphical querying
• CPC Class Codes: G06V 20/584 of vehicle lights or traffic lights
• CPC Class Codes: G08G 1/0175 by photographing vehicles, e.g. when violating traffic rules

Abstract

A vehicle model identification system includes an imaging device, a display device, a vehicle model identification device, and a bus that connects therebetween, and the vehicle model identification device includes an input device, a storage device, and a processor. A user performs a search with the input device, and the processor causes the display device to display a first vehicle image list, a vehicle image, and a reference image according to a first rule from the search condition. The user refers to the reference image, examines a feature of a target vehicle model, and specifies a partial region which is a feature. The processor acquires partial specification information of the specified partial region, searches for the target vehicle model based on a second rule, and displays a second vehicle image list and the vehicle image on the display device.

Claims

1. A vehicle model identification device for identifying a model of a vehicle based on a vehicle image of the vehicle captured by an imaging device, the device comprising:

a processor; and

a storage device,

wherein the storage device records the vehicle image and a score indicating a probability that the vehicle in the vehicle image is a specific vehicle model, and

the processor

acquires search condition including information for specifying a vehicle model,

extracts a first vehicle image list that matches the search condition by using a first rule based on the search condition and the score,

displays the first vehicle image list on a display device,

acquires partial specification information that specifies at least a part of the vehicle image,

generates a second rule based on the partial specification information, and

extracts a second vehicle image list that matches the search condition by using the generated second rule based on the search condition and the score.

2. The vehicle model identification device of claim 1,

wherein the score is defined for each of a plurality of partial regions in the vehicle image, and

the second rule is a rule of increasing a weight of the score of at least one partial region as compared with the first rule.

3. The vehicle model identification device of claim 2,

wherein the second rule uses the score of only the at least one partial region.

4. The vehicle model identification device of claim 2,

wherein the first rule is a rule of using a score obtained by evenly weighting all partial regions of a front image in the vehicle image, and

the second rule is a rule of increasing the weight of the score of at least one partial region included in the front image as compared to the weights of the scores of the other partial regions included in the front image.

5. The vehicle model identification device of claim 2,

wherein the score is calculated based on a feature amount of each partial region.

6. The vehicle model identification device of claim 5,

wherein the feature amount is a numerical value calculated by dense SIFT (Scale-Invariant Feature Transform).

7. The vehicle model identification device of claim 1,

wherein the processor uses the second rule to narrow down the first vehicle image list to the second vehicle image list having a smaller number of vehicle images.

8. The vehicle model identification device of claim 1,

wherein the processor displays a reference image obtained by imaging a vehicle of the same model as the vehicle of the vehicle image matching the search condition on the display device together with the first vehicle image list or the second vehicle image list.

9. A vehicle model identification system comprising:

the vehicle model identification device of claim 1;

the imaging device for imaging a vehicle; and

a display device that displays the vehicle image, the first vehicle image list, and the second vehicle image list;

10. A vehicle model identification method for identifying a model of a vehicle based on a vehicle image of the vehicle captured by an imaging device, the method comprising:

recording a vehicle image that is an image in which a vehicle is imaged and a score indicating the probability that the vehicle in the vehicle image is a specific vehicle model;

acquiring search condition that is information for specifying a vehicle model;

extracting a first vehicle image list that matches the search condition by using a first rule based on the search condition and the score;

displaying the first vehicle image list on a display device;

acquiring partial specification information that specifies at least a part of the vehicle image;

generating a second rule based on the partial specification information, and

extracting a second vehicle image list that matches the search condition by using the generated second rule based on the search condition and the score.