What is a point cloud ? 1. INTRODUCTION 2. OBJECTIVE 3. - - PDF document

DRIVER VISIBILITY ASSESMENT ON URBAN CROSSING ENVIRONMENTS BASED ON POINT CLOUDS

• L. Díaz-Vilariño, C. Silva, P. Arias

31th ICTCT Conference On the track of future urban mobility: safety, human factors and technology Porto, October 26th, 2018

INDEX

• 1. INTRODUCTION
• 2. OBJECTIVE
• 3. METHODOLOGY

3.1. Visibility analysis 3.2. 2D Isovist Field 3.3. 3D Isovist Field

• 4. EXPERIMENTS & RESULTS
• 5. CONCLUSIONS

INDEX

• 1. INTRODUCTION
• 2. OBJECTIVE
• 3. METHODOLOGY

3.1. Visibility analysis 3.2. 2D Isovist Field 3.3. 3D Isovist Field

• 4. EXPERIMENTS & RESULTS
• 5. CONCLUSIONS

What is a point cloud?

• 1. INTRODUCTION

What is a point cloud?

• Geometry
• Intensity
• Colour
• (…)
• 1. INTRODUCTION

What is a point cloud?

Principle of measurement Total Station vs Laser Scanner Measurement ‘point by point’ Measurement rate: 1,000,000 points/sec

• 1. INTRODUCTION

What is a point cloud?

Platforms

• Static
• Mobile
• Mobile low-weight
• 1. INTRODUCTION

What is a point cloud?

• Mobile
• 1. INTRODUCTION

Point clouds -Urban and interurban infrastructures Represent existing reality High quality

• Accuracy
• Resolution

Huge amount of (big) data Methodologies to automatically process data and extract useful information

What is a point cloud?

Autonomous vehicles

• 1. INTRODUCTION
• Perceiving the as-built environment
• Real-time applications (obstacle detection,

• Accuracy
• Resolution

Huge amount of (big) data Methodologies to automatically process data and extract useful information

• Low-quality data

What is a point cloud?

• 1. INTRODUCTION

Terrestrial Laser Scanners

• Analysing and understanding the as-

• Accuracy
• Resolution

Huge amount of (big) data Methodologies to automatically process data and extract useful information

• High-quality data

INDEX

• 1. INTRODUCTION
• 2. OBJECTIVE
• 3. METHODOLOGY

3.1. Visibility analysis 3.2. 2D Isovist Field 3.3. 3D Isovist Field

• 4. EXPERIMENTS & RESULTS
• 5. CONCLUSIONS

• 2. OBJECTIVE

To develop a methodology to the automatic analysis of visibility in urban crossings based on point clouds

• 2. OBJECTIVE

• Poor visibility is highlighted as an important cause of road accidents
• 3D realistic models of the as-built environment are essential for accurate analysis

To develop a methodology to the automatic analysis of visibility in urban crossings based on point clouds Why?

INDEX

• 1. INTRODUCTION
• 2. OBJECTIVE
• 3. METHODOLOGY

3.1. Visibility analysis 3.2. 2D Isovist Field 3.3. 3D Isovist Field

• 4. EXPERIMENTS & RESULTS
• 5. CONCLUSIONS

• 3. METHODOLOGY

2D Isovist 3D Isovist Ray-tracing algorithm Occlusion detection Point cloud GENERAL WORKFLOW

• 3. METHODOLOGY

3.1. VISIBILITY ANALYSIS Visibility analysis (2D or 3D) is performed for each position of an observer, considering:

• Visual angle
• Maximum line of sight
• Eye gaze direction

Ray-tracing algorithm to detect occlusions (obstacles interfering with LoSs) 3.1.1. Ray-tracing algorithm Scene is discretized in LoSs covering the field of view

• 3. METHODOLOGY

3.1. VISIBILITY ANALYSIS 3.1.2. Occlusion condition Existence of points intersecting with LoSs (within a buffer of radius r)

• Figure. Occlusion condition

Visibility analysis (2D or 3D) is performed for each position of an observer, considering:

• Visual angle
• Maximum line of sight
• Eye gaze direction

Ray-tracing algorithm to detect occlusions (obstacles interfering with LoSs)

• Isovist theory (1979) defines an isovist as a polygon representing the space seen form one
• bservation point

• 3. METHODOLOGY

3.2. 2D ISOVIST FIELD

• Figure. 2D human visual field.
• Isovist defined by occlusion points in 2D
• Figure. Isovist
• Space view defined by:
• visual angle (2D)
• maximum line of sight
• eye gaze direction

• 3. METHODOLOGY

• Figure. 3D human visual field.

3.3. 3D ISOVIST FIELD

• Extending the Isovist theory from 2D to 3D based on the direct use of point clouds.
• Isovist defined by occlusion points in 3D
• Space view defined by:
• visual angle (3D)
• maximum line of sight
• eye gaze direction

INDEX

• 1. INTRODUCTION
• 2. OBJECTIVE
• 3. METHODOLOGY

3.1. Visibility analysis 3.2. 2D Isovist Field 3.3. 3D Isovist Field

• 4. EXPERIMENTS & RESULTS
• 5. CONCLUSIONS

• 4. EXPERIMENTS & RESULTS

Case study 1 – Largo do Padrão (Porto)

• 6 hours
• 15,4 kms
• 33.000 106 points

• 4. EXPERIMENTS & RESULTS

Case study 1 – Largo do Padrão (Porto)

• 4. EXPERIMENTS & RESULTS

2D ISOVISTS

• Max. Line of Sight

• 30º to 30º

• 4. EXPERIMENTS & RESULTS

A B C

2D ISOVISTS

• 4. EXPERIMENTS & RESULTS

3D ISOVISTS

• Max. Line of Sight

• 30º to 30º

• Ang. Res.

• 4. EXPERIMENTS & RESULTS

3D ISOVISTS

INDEX

• 1. INTRODUCTION
• 2. OBJECTIVE
• 3. METHODOLOGY

3.1. Visibility analysis 3.2. 2D Isovist Field 3.3. 3D Isovist Field

• 4. EXPERIMENTS & RESULTS
• 5. CONCLUSIONS

• 5. CONCLUSIONS

• A methodology to automatically generate 2D and 3D isovists from point clouds in urban scenes.
• Visibility is studied for the space view defined by visual angle, maximum line of sight and eye

gaze direction.

• Results depend on data quality, specially in terms of completeness and noise.
• Results show how obstacles interfere in the visibility from a 3D perspective (understand the

perception and behaviour in urban scenes)

DRIVER VISIBILITY ASSESMENT ON URBAN CROSSING ENVIRONMENTS BASED ON POINT CLOUDS

• L. Díaz-Vilariño, C. Silva, P. Arias

31th ICTCT Conference On the track of future urban mobility: safety, human factors and technology Porto, October 26th, 2018 lucia@uvigo.es