Human detection apparatus and method using low-resolution two-dimensional (2D) light detection and ranging (LIDAR) sensor

Patent metadata

Field	Value
Publication number	US-10366282-B2
Application number	US-201715618877-A
Country	US
Kind code	B2
Filing date	Jun 9, 2017
Priority date	Jun 9, 2016
Publication date	Jul 30, 2019
Grant date	Jul 30, 2019

Abstract

A human detection apparatus and method using low-resolution two-dimensional (2D) light detection and ranging (LIDAR) sensor are provided. The human detection method may include receiving LIDAR data generated by reflecting a laser signal that continues to be transmitted to a search region from a plurality of objects in the search region, clustering a plurality of points included in the received LIDAR data by the same objects based on a correlation between the plurality of points, deriving a characteristic function used to identify a shape of a human, based on the clustered points, and determining whether each of the objects is a human based on the derived characteristic function.

First claim

What is claimed is:

1. A human detection method performed by a processor using a light detection and ranging (LIDAR) sensor, the human detection method comprising:
receiving LIDAR data generated by reflecting a laser signal that continues to be transmitted to a search region from a plurality of objects in the search region;
clustering a plurality of points included in the received LIDAR data by the same objects based on a correlation between the plurality of points;
applying the plurality of clustered points to a higher-order polynomial to derived a higher-order characteristic function, the higher-order characteristic function being used to identify a shape of a human; and
determining whether each of the objects is a human based on the derived higher-order characteristic function.

2. The human detection method of claim 1 , further comprising:
filtering invalid LIDAR data among the received LIDAR data,
wherein the clustering comprises clustering a plurality of points included in the filtered LIDAR data by the same objects based on a correlation between the plurality of points included in the filtered LIDAR data.

3. The human detection method of claim 1 , further comprising:
setting a region of interest (ROI) in which a human is to be sensed in the search region,
wherein the clustering comprises clustering a plurality of points included in LIDAR data corresponding to the set ROI by the same objects based on a correlation between the plurality of points included in the LIDAR data corresponding to the set ROI.

4. The human detection method of claim 1 , wherein the higher-order characteristic function is a multi-order polynomial curve function.

5. The human detection method of claim 1 , wherein the determining comprises:
extracting human feature data based on the derived higher-order characteristic function; and
comparing the extracted human feature data to prestored human feature data, to determine whether each of the objects is a human.

6. The human detection method of claim 5 , wherein the human feature data comprises at least one of information about a width, a circumference and a height of a human body.

7. A non-transitory computer readable recording medium storing computer readable instructions to cause at least one processor to implement the method of claim 1 .

8. A human detection apparatus using a light detection and ranging (LIDAR) sensor, the human detection apparatus comprising:
a processor, the processor comprising:
a receiver configured to receive LIDAR data generated by reflecting a laser signal that continues to be transmitted to a search region from a plurality of objects in the search region;
a clusterer configured to cluster a plurality of points included in the received LIDAR data by the same objects based on a correlation between the plurality of points;
a deriver configured to apply the clustered points to a higher-order polynomial to derive a higher-order characteristic function, the higher-order characteristic function being used to identify a shape of a human; and
a determiner configured to determine whether each of the objects is a human based on the derived higher-order characteristic function.

9. The human detection apparatus of claim 8 , further comprising:
a filter configured to filter invalid LIDAR data among the received LIDAR data,
wherein the clusterer is further configured to cluster a plurality of points included in the filtered LIDAR data by the same objects based on a correlation between the plurality of points included in the filtered LIDAR data.

10. The human detection apparatus of claim 8 , further comprising:
a setter configured to set a region of interest (ROI) in which a human is to be sensed in the search region,
wherein the clusterer is further configured to cluster a plurality of points included in LIDAR data corresponding to the set ROI by the same objects based on a correlation between the plurality of points included in the LIDAR data corresponding to the set ROI.

11. The human detection apparatus of claim 8 , wherein the determiner is further configured to extract human feature data based on the derived higher-order characteristic function, to compare the extracted human feature data to prestored human feature data, and to determine whether each of the objects is a human.

Assignees

• Daegu Gyeongbuk Inst Science & Tech

Inventors

• Lee Jong Hun
• Kwon Seong Kyung
• Son Sang Hyuk
• Hyun Eugin
• Lee Jin Hee

Classifications

• G01S17/93

  for anti-collision purposes · CPC title

• G01S17/89

  for mapping or imaging · CPC title

• G01S17/006

  Theoretical aspects · CPC title

• G01S7/4808

  Evaluating distance, position or velocity data · CPC title

• G06T7/11

  Region-based segmentation · CPC title