monitoring an intersection using a network of laser scanners

MONITORING AN INTERSECTION USING
A NETWORK OF LASER SCANNERS
H. Zhao, J. Cui, H. Zha,
Peking University
K. Katabira, X. Shao, R. Shibasaki,
University of Tokyo

Background (1)
Analyzing and Monitoring
the traffic behavior in an
intersection
Efficiently and accurately
COLLECTING the TRAFFIC
DATA in an INTERSECTION
Real-timely DETECTING
DANGEOUS SITUATIONS.

Background (2)
• Vision-based methods suffer mainly on the
following difficulties
–Occlusion
– Computation Cost
– Illumination Change
To solve the problems
e.g. the camera is
required to be set on a
tall position,
monitoring intersection
from the above
1. Restrict camera’s setting condition
2. Target on a simplified situation
e.g. monitor
vehicles of
limited lanes, do
not discriminate
moving objects.

Objective
This research propose a novel system for
monitoring and collecting detailed traffic data,
with easy setting condition, in an environment of
complicated traffic behavior, such as intersection,
using a network of single-row laser scanner.

Pedes.
Car
#1
System Image
Server PC
Bicycle
#2
Car
An image of integrated laser data
Network
#1
Laser Scanner
#2
Laser Scanner

Road
side
LD3
LD1
Road
side
LD1
Safety
zone
Safety
zone
LD2
Integrated frame
Single laser frame
LD3
10m
Single laser frame
LD2
Single laser frame

Client
Processing Flow
Laser Scanner1
Measure data
Laser Scanner N
Measure data
BG Generation
BG Subtraction
…
BG Generation
BG Subtraction
Clustering
Clustering
Network
Server
1. Data Integration
2. Detection
3. Tracking
4. Recognition
Moving object

Processing Modules
Server
Network
Recognition
Tracking
Detection
Data Integration
Clustering
Moving object
Calibration, Synchronization
Clustering
BG Generation
BG Subtraction
BG Generation
BG Subtraction
Client
Measure Data
Laser Scanner 1
…
Measure Data
Laser Scanner N

Server
Recognition
Frame k
Tracking
Detection
Data Integration
trajectory
state
group
observation
Network
Clustering
cluster
BG Generation
BG Subtraction
Client
Measure Data
LD1 LD2 LD3
Data Structure

Difficulties

Difficulty 1

The data of the same object might be spatially disconnected

Difficulties

The data of different objects might be spatially close
There are different kinds of objects, complicated motions

Task 1
We are not able to rely only on distance,
size, shape, location etc for detection
and traction.
We need to find some other clues to
associate the data of the same object
together.

Data Feature
A car
s
e
u1
c
u2

Data Feature
v4
u
u
u
u
v3
v1
u
u
u
v2

Data Feature
u
u
u
u
u
u

Object Model
(c, val)
(v3, sv3)
u
(v4, sv4)
p
(v2, sv2)
dirv
u
len2
(v1, sv1)
len1
(v1=dirv)

Difficulty 2
The current
frame
The previous
frame
More sensors bring more confusion in data association

Task 2
Data association is difficult when facing a
complicated environment such as an
intersection, where different kinds of objects,
different motion patterns exist.
Also, a large number of distributed sensors will
bring more difficulties in data association.
A data association algorithm is required to tackle
the confusions.

A range stream
Scan angle (within a laser scan)
a static object
a moving object
Time (A stream of laser scans)

A simple clustering result
Time (A stream of laser scans)
Scan angle (within a laser scan)

Data Association Method
k-1 k
trajectory
New trajectory
3
group
2
cluster
1
LD1 LD2 LD3 LD1 LD2 LD3 LD3

Experiment
Laser Scanner
Video Camera
Sports Center
Peking Univ.
Overhead Bridge
Peking Univ.

40cm
激 光 扫 描 仪

Summary
We propose a novel system for monitoring and collecting
detailed traffic data, with easy setting condition, in an
environment of complicated traffic behavior, such as
intersection, using a network of single-row laser scanners.
Future Topics
• Accuracy evaluation
• Recognition module
• Fusing laser scanner with video camera