# US 2009/0024251 A1 · Method and apparatus for estimating pose of mobile robot using particle filter

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

- Number: US 2009/0024251 A1
- Document ID: US20090024251A1
- Status: Published application
- Kind: A1
- Application number: 12/155,552
- Filed: 2008-06-05
- Published: 2009-01-22
- Earliest claimed: 2007-07-18
- Assignees: Samsung Electronics Co., Ltd.
- Inventors: Hyeon Myeong; Woo-Yeon Jeong; Seok-Won Bang
- Web page: https://patentgrep.com/patent/US20090024251A1

## CPC

- G05D 1/0246: {using a video camera in combination with image processing means}
- G05D 1/0272: {comprising means for registering the travel distance, e.g. revolutions of wheels (measuring distance traversed on the ground by vehicles, e.g. using odometers G01C22/00)}
- G05D 1/0274: {using mapping information stored in a memory device (navigation using map-matching G01C21/30)}
- G05D: SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES

## Abstract

A method and apparatus for estimating the pose of a mobile robot using a particle filter is provided. The apparatus includes an odometer which detects a variation in the pose of a mobile robot, a feature-processing module which extracts at least one feature from an upward image captured by the mobile robot, and a particle filter module which determines current poses and weights of a plurality of particles by applying the mobile robot pose variation detected by the odometer and the feature extracted by the feature-processing module to previous poses and weights of the particles.

## Claims

### Claim 1

An apparatus for estimating a pose of a mobile robot, comprising:
an odometer which detects a variation in the pose of a mobile robot;
a feature-processing module which extracts at least one feature from an upward image captured by the mobile robot; and
a particle filter module which determines current poses and weights of a plurality of particles by applying the mobile robot pose variation detected by the odometer and the feature extracted by the feature-processing module to previous poses and weights of the particles.

### Claim 2 (depends on 1)

The apparatus of claim 1, wherein the feature-processing module comprises:
a feature-extraction unit which extracts a current feature from an image captured at a current time by the mobile robot; and
a feature-matching unit which determines whether the current feature is identical to a previous feature extracted at a previous time.

### Claim 3 (depends on 2)

The apparatus of claim 2, wherein the feature-matching unit determines whether the current feature is identical to the previous feature by sharing a local image including the feature extracted by the feature-processing module for each the particle.

### Claim 4 (depends on 3)

The apparatus of claim 3, wherein the feature-matching unit comprises a data structure which includes coordinates of a feature and a local image including the feature.

### Claim 5 (depends on 1)

The apparatus of claim 1, wherein the feature extracted by the feature-processing module is a feature point or a line with multiple feature points.

### Claim 6 (depends on 1)

The apparatus of claim 1, wherein the feature extracted by the feature-processing module is a line or a point extracted from a contour obtained using the Maximally Stable Extremal Region (MSER) technique.

### Claim 7 (depends on 1)

The apparatus of claim 1, wherein the particle filter module comprises:
a pose-estimation unit which determines the current poses of the particles by applying the mobile robot pose variation detected by the odometer to the particles; and
a weight-calculation unit which compares the current poses of the particles with a pose of the mobile robot estimated based on the feature extracted by the feature-processing module and allocates high weights to particles whose current poses are similar to the estimated mobile robot pose.

### Claim 8 (depends on 7)

The apparatus of claim 7, wherein the particle filter module further comprises a sampling unit which resamples a number of particles according to the weights of the particles.

### Claim 9 (depends on 7)

The apparatus of claim 7, wherein the pose-estimation unit calculates a variation in the pose of a particle by adding up the mobile robot pose variation detected by the odometer and random noise within a noise covariance range.

### Claim 10 (depends on 8)

The apparatus of claim 8, wherein the pose-estimation unit increases covariance if a difference between the mobile robot pose variation detected by the odometer and a variation in the pose of the mobile robot estimated based on the feature extracted by the feature-processing module is greater than a predefined threshold.

### Claim 11 (depends on 8)

The apparatus of claim 8, wherein the pose-estimation unit increases straight advance noise covariance if it is determined that the mobile robot has slipped on a floor and a difference between a variation in the pose of the mobile robot detected by a gyroscope and a variation in the pose of the mobile robot detected by an encoder is less than a predefined threshold; and increases rotation noise covariance if it is determined that the mobile robot has slipped on the floor and the difference between the mobile robot pose variation detected by the gyroscope and the mobile robot pose variation detected by the encoder is greater than the predefined threshold.

### Claim 12 (depends on 1)

The apparatus of claim 1, wherein the odometer comprises at least one of an encoder and a gyroscope.

### Claim 13

A method of estimating a pose of a mobile robot, comprising:
(a) detecting a variation in the pose of a mobile robot;
(b) extracting at least one feature from an upward image captured by the mobile robot; and
(c) determining current poses and weights of particles by applying the detected mobile robot pose variation and the extracted feature to previous poses and weights of the particles.

### Claim 14 (depends on 13)

The method of claim 13, wherein (b) comprises:
(b1) extracting a current feature from an image captured at a current time by the mobile robot; and
(b2) determining whether the current feature is identical to a previous feature extracted at a previous time.

### Claim 15 (depends on 14)

The method of claim 14, wherein (b2) comprises sharing a local image including the extracted feature between the particles.

### Claim 16 (depends on 13)

The method of claim 13, wherein the extracted feature is a line or a point extracted from a contour obtained using the MSER technique.

### Claim 17 (depends on 13)

The method of claim 13, wherein (c) comprises
(c1) determining the current poses of the particles by applying the detected mobile robot pose variation to the particles; and
(c2) comparing the current poses of the particles with a pose of the mobile robot estimated based on the extracted feature, and allocating high weights to particles whose current poses are similar to the estimated mobile robot pose.

### Claim 18 (depends on 17)

The method of claim 17, wherein (c1) comprises calculating a variation in the pose of a particle by adding up the detected mobile robot pose variation and random noise within a noise covariance range.

### Claim 19 (depends on 18)

The method of claim 18, wherein (c1) further comprises increasing covariance if a difference between the detected mobile robot pose variation and the estimated mobile robot pose is greater than a predefined threshold.

### Claim 20 (depends on 18)

The method of claim 18, wherein (c1) further comprises, if it is determined that the mobile robot has slipped on a floor:
increasing straight advance noise covariance if a difference between a variation in the pose of the mobile robot detected by a gyroscope and a variation in the pose of the mobile robot detected by an encoder is less than a predefined threshold; and
increasing rotation noise covariance if the difference between the mobile robot pose variation detected by the gyroscope and the mobile robot pose variation detected by the encoder is greater than the predefined threshold.
