Detection device, learning device, detection method, learning method, and information storage device

Patent metadata

Field	Value
Publication number	US-9454711-B2
Application number	US-201414290006-A
Country	US
Kind code	B2
Filing date	May 29, 2014
Priority date	May 30, 2013
Publication date	Sep 27, 2016
Grant date	Sep 27, 2016

Abstract

A detection device includes an image acquisition section that acquires an image that has been captured by an imaging section, and includes an image of an object, a distance information acquisition section that acquires distance information based on a distance from the imaging section to the object when the imaging section has captured the image, a feature quantity calculation section that calculates a feature quantity from the acquired image, the feature quantity relating to at least one of a color, a brightness, a color difference, and a spectrum of the object, a learning feature quantity storage section that stores a learning feature quantity calculated by a learning process based on the distance from the imaging section to the object, and a detection section that detects a target area from the image based on the learning feature quantity, the distance information, and the feature quantity.

First claim

What is claimed is:

1. A detection device comprising:
a processor comprising hardware, wherein the processor is configured to:
acquire an image of a luminal object, wherein the image has been captured by an image sensor;
acquire distance information on a distance from the image sensor to the luminal object at the time the image sensor captured the image;
calculate a feature quantity from the image, wherein the feature quantity relates to at least one of a color, a brightness, a color difference, and a spectrum of the luminal object;
perform a correction process on the feature quantity using structural information about the luminal object extracted from the image based on the distance information;
acquire a learning feature quantity stored in a learning feature quantity storage, wherein the learning feature quantity is calculated by a learning process performed on the feature quantity that has been subjected to the correction process using the structural information about the luminal object extracted from the image based on the distance information; and
detect a target area from the image based on the learning feature quantity and the feature quantity that has been subjected to the correction process using the structural information about the luminal object extracted from the image based on the distance information;
wherein the structural information is information about irregular structures of the luminal object.

2. The detection device as defined in claim 1 , wherein the processor is further configured to:
calculate the learning feature quantity by performing the learning process on the feature quantity that has been subjected to the correction process using the structural information about the luminal object extracted from the image based on the distance information.

3. The detection device as defined in claim 1 ,
wherein the processor is further configured to:
perform a highlight process on the image to highlight the target area.

4. The detection device as defined in claim 1 , wherein the processor is further configured to:
acquire the structural information by acquiring three-dimensional structural information about the luminal object from the distance information.

5. A learning device comprising:
a processor comprising hardware, wherein the processor is configured to:
acquire an image of a luminal object, wherein the image has been captured by an image sensor;
acquire distance information on a distance from the imaging sensor to the luminal object at the time the image sensor captured the image;
calculate a feature quantity from the image, wherein the feature quantity relates to at least one of a color, a brightness, a color difference, and a spectrum of the luminal object;
perform a correction process on the feature quantity using structural information about the luminal object extracted from the image based on the distance information; acquire a learning feature quantity by performing a learning process on the feature quantity that has been subjected to the correction process; and
controlling a learning feature quantity storage to store the learning feature quantity calculated by the learning process performed on the feature quantity that has been subjected to the correction process;
wherein the structural information is information about irregular structures of the luminal object.

6. The learning device as defined in claim 5 , wherein the processor is further configured to:
acquire the structural information by acquiring three-dimensional structural information about the luminal object from the distance information.

7. A detection method comprising causing a computer to perform:
acquiring an image of a luminal object, wherein the image has been captured by an image sensor;
acquiring distance information on a distance from the image sensor to the luminal object at the time the image sensor captured the image;
calculating a feature quantity from the image, wherein the feature quantity relates to at least one of a color, a brightness, a color difference, and a spectrum of the luminal object;
performing a correction process on the feature quantity using structural information about the luminal object extracted from the image based on the distance information;
acquiring a learning feature quantity stored in a learning feature quantity storage, wherein the learning feature quantity is calculated by a learning process performed on the feature quantity that has been subjected to the correction process using the structural information about the luminal object extracted from the image based on the distance information; and
detecting a target area from the image based on the learning feature quantity and the feature quantity that has been subjected to the correction process using the structural information about the luminal object extracted from the image based on the distance information;
wherein the structural information is information about irregular structures of the luminal object.

8. A learning method comprising causing a computer to perform:
acquiring an image of a luminal object, wherein the image has been captured by an image sensor;
acquire distance information on a distance from the image sensor to the object at the time the image sensor captured the image;
calculate a feature quantity from the image, wherein the feature quantity relates to at least one of a color, a brightness, a color difference, and a spectrum of the luminal object;
perform a correction process on the feature quantity using structural information about the luminal object extracted from the image based on the distance information;
acquire a learning feature quantity by performing a learning process on the feature quantity that has been subjected to the correction process; and
controlling a learning feature quantity storage to store the learning feature quantity acquired by the learning process performed on the feature quantity that has been subjected to the correction process;
wherein the structural information is information about irregular structures of the luminal object.

9. A computer-readable storage device with an executable program stored thereon, wherein the executable program instructs a computer to perform:
acquiring an image of a luminal object, wherein the image has been captured by an image sensor;
acquiring distance information on a distance from the image sensor to the luminal object at the time the image sensor captured the image;
calculating a feature quantity from the image, wherein the feature quantity relates to at least one of a color, a brightness, a color difference, and a spectrum of the luminal object;
performing a correction process on the feature quantity using structural information about the luminal object extracted from the image based on the distance information;
acquiring a learning feature quantity stored in a learning feature quantity storage, wherein the learning feature quantity is calculated by a learning process performed on the feature quantity that has been subjected to the correction process using the structural information about the luminal object extracted from the image based on the distance information; and
detecting a target area from the image based on the learning feature quantity and the feature quantity that has been subjected to the correction process using the structural information about the luminal object extracted from the image based on the distance information;
wherein the structural information is information about irregular structures of the luminal object.

10. A computer-readable storage device with an executable program stored thereon, wherein the executable program instructs a computer to perform:
acquiring an image of a lum

Assignees

• Olympus Corp

Inventors

• Mitsui Masanori

Classifications

• G06V10/811

  the classifiers operating on different input data, e.g. multi-modal recognition · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06V20/698 Primary

  Matching; Classification · CPC title

• G06V10/56

  relating to colour · CPC title

• G06F18/256

  of results relating to different input data, e.g. multimodal recognition · CPC title