Apparatus, method for detecting critical areas and pedestrian detection apparatus using the same

• US 8,164,432 B2
• Filed: 05/26/2009
• Issued: 04/24/2012
• Est. Priority Date: 06/04/2008
• Status: Active Grant

• Current Assignee: HL Klemove Corporation
• Original Assignee: Mando Corporation
• Inventors: Broggi, Alberto, Cerri, Pietro, Jung, Ho-gi
• Primary Examiner(s): PHAM, TOAN NGOC
• Application Number:
  US12/471,878
• Publication Number:
  US 20090303026A1
• Time in Patent Office: 1,064 Days
• Field of Search: 340/435, 340/425.5, 340/903, 340/436, 340/944, 382/103, 382/104, 348/148, 348/149
• US Class Current: 340/435
• CPC Class Codes: G01S 17/86 Combinations of lidar systems with systems other than lidar, radar or sonar, e.g. with direction finders
• CPC Class Codes: G01S 17/931 of land vehicles
• CPC Class Codes: G06V 20/58 Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
• CPC Class Codes: G06V 20/64 Three-dimensional objects
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition
• CPC Class Codes: G08G 1/165 for passive traffic, e.g. including static obstacles, trees
• CPC Class Codes: G08G 1/166 for active traffic, e.g. moving vehicles, pedestrians, bikes
• CPC Class Codes: G08G 1/167 Driving aids for lane monitoring, lane changing, e.g. blind spot detection

Abstract

An apparatus, method for detecting critical areas and a pedestrian detection apparatus using the same are provided. An application of the pedestrian detection system is provided to help limit critical urban environment to particular areas. Contrary to traditional pedestrian detection systems that localize every pedestrians appearing in front of the subject vehicle, the apparatus first finds critical areas from urban environment and performs a focused search of pedestrians. The environment is reconstructed using a standard laser scanner but the subsequent checking for the presence of pedestrians is performed by incorporating a vision system. The apparatus identifies pedestrians within substantially limited image areas and results in boosts of timing performance, since no evaluation of critical degrees is necessary until an actual pedestrian is informed to the driver or onboard computer.

Claims

1. A pedestrian detection apparatus for detecting pedestrians about a vehicle comprising:

an image generator for capturing images of an environment of the vehicle to generate and deliver image outputs;

a sensor for sensing obstacles about the vehicle and delivering distance data between the obstacles and the vehicle;

a critical area detector for generating polylines by clustering a plurality of the distance data and classifying obtained polylines according to their size and shape into multiple categories in order to detect multiple categories of obstacles; and

a pedestrian detector for detecting the pedestrians preferentially from images corresponding to the critical areas.

2. The pedestrian detection apparatus in claim 1, wherein the image generator comprises a near infrared camera and a plurality of near infrared LED headlights.

3. The pedestrian detection apparatus in claim 1, wherein the pedestrian detector preferentially detects pedestrians from the images that correspond to the critical areas only upon occurrence of the moving obstacles during a detection of pedestrians from the images or selected candidate images.

4. A critical area detector in a pedestrian detection apparatus for detecting pedestrians about vehicles comprising:

a critical area detection unit, which upon receiving a distance data generated from detecting environmental obstacles of the vehicles, generates polylines by clustering a plurality of the distance data and classifies obtained polylines according to their size and shape into multiple categories in order to detect multiple categories of obstacles.

5. The critical area detector in claim 4, receiving the distance data from a sensor, which is equipped with an interlaced scanning laser scanner.

6. The critical area detector in claim 4, wherein the critical area detection unit comprises:

a data correction section to receive the distance data from a sensor and correct the same according to the positional changes of the vehicles;

a data clustering section for connecting between the corrected distance data in a close proximity and clustering them as chains of segments;

a segment merging section for connecting the clustered 5 distance data segments to merge adjacent segments with approximately the same orientation into a polyline;

a line merging section for integrating the polylines in a close proximity into a single merged line;

a second obstacle classification section for correcting the polylines according to the vehicle positional changes and classifying the polylines into multiple categories according to the degree of overlap between the polylines before correction and the polylines after correction in order to detect multiple categories of obstacles; and

a critical area identification section for detecting the critical areas by identifying the environment and the moving obstacles with using the multiple categories of obstacles detected in the first and second obstacle classification sections.

7. The critical area detector in claim 6, wherein the correction section corrects the distance data by calculating a vehicle roto-translation using ego-motion data received from a vehicle CAN.

8. The critical area detector in claim 6, wherein the data clustering section clusters the distance data by connecting among continuous pulses of the distance data only the ones that belong to the same rotation.

9. The critical area detector in claim 6, wherein the first obstacle classification section detects one or more of multiple categories of obstacles including a possible pedestrian, road border, L-shaped obstacle, moving obstacle, and generic obstacle.

10. The critical area detector in claim 6, wherein the second obstacle classification section generates the corrected polylines by moving the polylines according to ego-motion.

11. The critical area detector in claim 6, wherein the second obstacle classification section detects one or more of multiple categories of obstacles including a moving obstacle, static obstacle, changing shape obstacle, and new obstacle.

12. A critical area detection method in a pedestrian detection apparatus for detecting pedestrians about a vehicle comprising:

(a) correcting a distance data upon receiving the same from a sensor according to positional changes of the vehicle;

(b) clustering the distance data as chains of segments by connecting between corrected distance data in a close proximity;

(c) merging the segments by connecting the clustered distance data segments and merging the segments in parallel and in a close proximity into a polyline;

(d) merging lines by integrating polylines in a close proximity into a single merged line;

(e) classifying obstacles through classification of polylines according to their size and shape into multiple categories in order to detect multiple categories of obstacles;

(f) classifying the obstacles through correcting the polylines according to the vehicle positional changes and classifying the polylines into multiple categories according to the degree of overlap between the polylines before correction and the polylines after correction in order to detect multiple categories of obstacles; and

(g) identifying critical areas through identifying the environment and the moving obstacles by using the multiple categories of obstacles detected in steps (e) and (f) of classifying the obstacles in order to detect the critical areas.