Synopsis

​

2-D vision systems are able to provide high-speed, consistent and accurate product inspection in the industry. However, they are not capable of detecting defects that manifest as height variation on the surface of the products, such as dents and protrusions. Neither are they able to measure heights and 3-D surface profiles of products. In this two-day training workshop, the participants will learn about the fundamentals of 3-D machine vision through intensive guided activities and hands-on practical. The most commonly used 3-D methods in machine vision will be covered in this training. These include stereo imaging, photogrammetry, laser displacement scanning, triangulation, structured lighting method, phase-shift fringe projection method, shadow moiré topography, and phase-measuring deflectometry method. Participants will be guided into writing OpenCV-Python codes for reconstructing 3-D surface data from phase-shifted fringe pattern images.

 

The training workshop is divided into activity-based theory (50%) to strengthen the fundamental knowledge of the participants on the underlying technologies of 3-D surface measurement, and hands-on practical activities (50%). The practical activities involve capturing 2-D images of smooth surfaces using special lighting methods and reconstructing 3-D surface data from the 2-D images. Participants may bring their own specimens for testing.

​

Course outcome

​

• Understand the basics of 3-D stereo imaging and photogrammetry

• Apply the laser displacement method to determine 3-D profile of a surface

• Understand the basics of the structured lighting method for 3-D surface measurement

• Generate phase-shifted fringe patterns in three- and four-shifts

• Write algorithms to obtained wrapped and unwrapped phase maps

• Write algorithm to reconstruct 3-D height data from unwrapped phase map

• Apply the phase-shift fringe projection method for 3-D surface profiling

• Apply the shadow moiré method to measure 3-D surface profile

• Apply the phase-measuring deflectometry method for measuring reflective surfaces

​

Course outline

​

Day 1

​

• Principles of 3-D stereo imaging & photogrammetry

• Laser displacement scanning & triangulation

• Structured lighting method

• Phase-shift fringe projection method

• Three- and four-shift methods

• Color fringe projection for 3-D profiling

• Phase unwrapping and 3-D surface reconstruction

• Activity 1 - Create stereoscopic image using mobile phone camera. (see activity sheet)

• Activity 2 - Create 3-D image using photogrammetry. (see activity sheet)

• Activity 3 (P) - Measure 3-D surface using laser displacement scanning. (see activity sheet)

• Activity 4 - Generate fringe pattern using OpenCV-Python. (see activity sheet)

• Activity 5 - Write OpenCV-Python code to generate phase-shifted fringe patterns. (see activity sheet)

• Activity 6 - Write OpenCV-Python code to generate phase map. (see activity sheet)

• Activity 7 - Write OpenCV-Python code to plot profile across phase map. (see activity sheet)

• Activity 8 - Write OpenCV-Python code to unwrap a phase map. (see activity sheet)

• Activity 9 - Write OpenCV-Python code to plot wrapped phase in 3-D. (see activity sheet)

• Activity 10 - Generate phase-shifted fringe patterns with ‘object’. (see activity sheet)

• Activity 11 - Generate phase map for fringe pattern with ‘object’. (see activity sheet)

• Activity 12 - Unwrap phase map with ‘object’. (see activity sheet) (see activity sheet)

• Activity 13a - Combine codes for unwrapping phase maps. (see activity sheet)

• Activity 13b - Subtract the two phase maps. (see activity sheet)

• Activity 14 (P) - Measure 3-D profile of shell using phase-shift fringe projection (see activity sheet)

​

Day 2

• Principles of the moiré method

• Shadow, projection and geometric moiré methods and their applications

• Phase-shift shadow moiré (three and four-shift methods) for warpage measurement

• 3-D surface reconstruction from moiré patterns

• Principles of deflectometry for specular surfaces

• Phase-measuring deflectometry method

• Activity 15 (P) - Create moiré fringe pattern using mobile phone camera. (see activity sheet)

• Activity 16 - Estimate height difference between points from moiré fringe pattern. (see activity sheet)

• Activity 17 - Write OpenCV-Python code to unwrap in two dimensions. (see activity sheet)

• Activity 18(P) - Measure 3-D profile of egg using phase-shift shadow moiré. (see activity sheet)

• Activity 19 - Determine slope data for curved mirror using phase-measuring deflectometry. (see activity sheet)

 

Who should attend:

Machine vision solution providers, engineers, and others interested to learn how to obtain 3-D data from 2-D images for integration into machine vision systems. You may assess yourself whether you need to take this training by answering the questions in this simple quiz.

​

Fee:

Please contract trainer for quotation.

​

Download sample slides here!

​

​Hardware available for hands-on activities: 6MP color camera, lenses, mini projector, translation stages

 

Sofware available: OpenCV & Python (Participants need to bring their own laptops - MS Windows only)