Robot cleaner and method for controlling the same

Patent metadata

Field	Value
Publication number	US-9375119-B2
Application number	US-201313785782-A
Country	US
Kind code	B2
Filing date	Mar 5, 2013
Priority date	Mar 9, 2012
Publication date	Jun 28, 2016
Grant date	Jun 28, 2016

Abstract

Disclosed are a robot cleaner, and a method for controlling the same.
Firstly, a sound source direction and a sound source position may be detected by one or more microphones, e,g., three microphones, and a specific event sound may be recognized. Then, the recognized specific event sound may be output to a cleaner body in the form of a message, or may be provided via a communication network. This can allow the circumstances to be easily monitored.
Secondly, a sound recognition function may be updated using an ambient sound, for sound recognition from background noise and environmental noise, in a robust and precise manner. This can enhance a sound recognition rate, and improve stability and efficiency.
Thirdly, the robot cleaner may be moved to a detected direction or position of a sound source with respect to an event sound. Then, image information may be detected, or whether an abnormal situation has occurred or not may be determined. Such information may be provided via a communication network.

First claim

What is claimed is:

1. A robot cleaner, comprising:
a sound input unit having one or more microphones, and configured to receive an ambient sound;
a sound recognition unit configured to recognize the received ambient sound, in a separation manner into an event sound and an environmental sound;
a message output unit configured to output a message corresponding to the event sound or the environmental sound;
a storage unit configured to store an event sound model with respect to the event sound, and an environmental sound model with respect to the environmental sound;
a control unit configured to detect a direction or a position of a sound source with respect to the event sound;
an image detection unit configured to detect image information by capturing the periphery of the sound source position; and
a communication unit configured to transmit, via a communication network, the message, or the image information, or both of the message and the image information,
wherein the sound recognition unit is configured to:
recognize a sound of the ambient sound, which matches the event sound model, as the event sound, the ambient sound being recognized at intervals of first time; and
recognize a sound of the ambient sound, which matches the environmental sound model, as the environmental sound, the ambient sound being recognized at intervals of second time longer than the first time.

2. The robot cleaner of claim 1 , wherein the sound input unit includes at least three microphones, and
wherein said at least three microphones are disposed inside or outside the robot cleaner with constant distances from each other.

3. The robot cleaner of claim 1 , wherein the sound recognition unit includes:
a first recognition portion configured to recognize the event sound at intervals of first time, based on the ambient sound and the event sound model;
a second recognition portion configured to recognize the environmental sound at intervals of first time, based on the ambient sound and the environmental sound model; and
a third recognition portion configured to recognize the environmental sound at intervals of second time longer than the first time, based on the ambient sound and the environmental sound model.

4. The robot cleaner of claim 3 , wherein the sound recognition unit updates the event sound model based on the environmental sound recognized by the third recognition portion.

5. The robot cleaner of claim 3 , wherein the sound recognition unit further includes a reliability determination portion configured to determine reliability on a recognition result acquired by the first recognition portion, based on a recognition result acquired by the second recognition portion, as well as a recognition result acquired by the first recognition portion.

6. The robot cleaner of claim 5 , wherein the reliability determination portion compares a likelihood of a first event sound having a largest likelihood among recognition results acquired by the first recognition portion, with a likelihood of a first environmental sound having a largest likelihood among recognition results acquired by the second recognition portion, and
wherein if the likelihood of the first event sound is greater than that of the first environmental sound as a result of the comparison, the reliability determination portion outputs the first event sound.

7. The robot cleaner of claim 6 , wherein if a difference between the first event sound having a largest likelihood and a second event sound having a second largest likelihood is more than a reference value, the reliability determination portion outputs the first event sound.

8. The robot cleaner of claim 3 , wherein the sound recognition unit further includes a feature extraction portion configured to extract feature vectors, from the ambient sound, based on a sound processing technique, and configured to output the extracted feature vectors to the first to third recognition portions.

9. A method for controlling a robot cleaner, comprising:
receiving an ambient sound based on one or more microphones;
recognizing the received ambient sound, by a sound recognition unit, in a separation manner into an event sound and an environmental sound;
outputting a message corresponding to the event sound or the environmental sound,
detecting, using a control unit, a direction or a position of a sound source with respect to the event sound;
detecting, using an image detection unit, image information by capturing the periphery of the sound source position; and
transmitting, using a communication unit, via a communication network, the message, or the image information, or both of the message and the image information,
wherein the step of recognizing the received ambient sound includes:
recognize a sound of the ambient sound which matches the event sound model, as the event sound, the ambient sound being recognized at intervals of first time; and
recognize a sound of the ambient sound, which matches the environmental sound model, as the environmental sound, the ambient sound being recognized at intervals of second time longer than the first time.

10. The method of claim 9 , further comprising:
moving a robot cleaner to the sound source position.

11. The method of one of claim 9 , the step of recognizing the received ambient sound includes:
extracting feature vectors from the ambient sound, based on a sound processing technique;
a first recognition process for recognizing the event sound at intervals of first time, based on the feature vectors and an event sound model with respect to the event sound;
a second recognition process for recognizing the environmental sound at intervals of first time, based on the feature vectors and an environmental sound model with respect to the environmental sound;
a third recognition process for recognizing the environmental sound at intervals of second time longer than the first time, based on the feature vectors and the environmental sound model; and
determining reliability on the recognized event sound based on a recognition result acquired in the second recognition process, as well as a recognition result acquired in the first recognition process.

12. The method of claim 11 , wherein the step of recognizing the received ambient sound further includes updating the event sound model based on the environmental sound recognized in the third recognition process.

Assignees

• Lg Electronics Inc

Inventors

• Lee Heungkyu
• Jeon Hyejeong
• Lee Byounggi

Classifications

• B25J9/0003

  Home robots, i.e. small robots for domestic use · CPC title

• Y10S901/01

  Mobile robot · CPC title

• G10L25/27

  characterised by the analysis technique · CPC title

• A47L9/009 Primary

  Carrying-vehicles; Arrangements of trollies or wheels; Means for avoiding mechanical obstacles · CPC title

• G08B13/19695

  Arrangements wherein non-video detectors start video recording or forwarding but do not generate an alarm themselves · CPC title