Reverse

Engineering
 Outline

Definition

Digitizing Methods

3D Reconstruction
 Definition
Reverse Engineering is the practice of studying a finished product, a
physical device or a software package, in order to learn something
about how it works or how it was manufactured. We shall be concerned
with RE of physical objects only.
Traditionally, RE is an accepted, legitimate practice, both in the
competitive marketplace and in technical education. It is comparable to
literature survey. It becomes unethical only when it is launched as a
product in competition to the original product it is illegal only if it is
prohibited in the form of patent or copyright which is zone-specific.
(Geometric) Reverse Engineering (RE) is the process of creating a
mathematical representation or CAD model of an object from its
physical form. Identification of the following are equally important:
- Tolerances and surface finish
- Reconstructing missing parts
- Material details (composition and condition)
- Process followed.
 Need for Reverse Engineering
A part is first modeled in clay, wood or foam by the stylist and
needs to be transferred into a CAD model.
Only 2D drawings or master models of the tools exist.
A change has been made into a physical part or tool. The CAD
model should be updated with this change.
Final parts have to be verified against the original CAD design.
An old equipment or a vital military hardware has broken down.
The supplier no longer exists or unwilling to supply the spare or it
is too expensive or it takes too long to import.
A patients left bone is damaged and an implant is required to be
prepared from the right side bone. There are many such medical
applications where RE, RP and RT play vital roles.
A competitor's product needs to be analyzed (copied !).
Stages of design methodology

Video
 Phases of Reverse Engineering
Geometry

i. Digitizing: Acquiring the point data

- This is known as digitizing.
- Several types of hardware are used to acquire data

ii. Reconstructing the 3D CAD model

- This is processing of the data using a software.

iii. Reconstruction of the missing parts

iv. Tolerances and fits

Material

Process
Digitizing
 Sensors

Touch probe
Laser scanning
Camera systems

Abcdh
8
 Touch Probe Measurement
The probes are available in different orientations
and end shapes so that different features in
various orientations can be measured in the
same set up.

The required probe is chosen or activated.

The probe often has a calibrated spherical end.

Calibration is required after every physical
change of probe or a collision. For this purpose,
special calibration spheres are available.

The load at which the probe must press the part

can be set. As the probe presses the part, when
the load exceeds the pre-set limit, it signals the
CMM to stop. It also passes on the force vector
from which the CMM calculates the coordinates
of the contact point.
 Laser Scanners

+ Non-contact - Less accurate

+ No collision - Only visible surfaces
+ Fast data capture
+ Soft material
Laser Scanners - Principle

Triangulation method:

An emitter and a receptor are

integral with the measuring
head at a constant distance.
These two, together with the
point of measurement the
point where the emitted pulse
of ray falls form a triangle.

When adequate level of signal

is received, the point on the
surface can be calculated
from the current position of
the measuring head and time
taken for the flight of a pulse
from the emitter to receptor.
 Camera Systems : Photogrammetry

One or more views of the object are captured as an

image on film or on an electronic image sensor. The
points are marked and referenced at distinct features
on the object (a,b,c).
The software calculates the position of the camera
for each photo.
It calculates the intersection of light rays from each of
the photo positions out into 3D space.
By using multiple photographs, you can capture the
whole object or scene.
 Camera Systems : Photogrammetry
Take pictures using digital, film, or
video cameras, and load them in to
PhotoModeler. Shoot two or more
overlapping photos from different
angles and then import them into the
program.

Next, mark and reference

features on the photographs
using PhotoModeler's Point,
Line and Edge tools. Using the
referencing functions, instruct
PhotoModeler by matching up
points across the photos.
 Camera Systems : Photogrammetry
Now PhotoModeler is ready to process
the camera and referencing data.
Using the "Process" menu,
PhotoModeler adjusts the input data
and creates 3D point data to produce
an accurate 3D model.

One can view, zoom, rotate, or measure

the 3D model in PhotoModeler's 3D
Viewer. The Point Table allows one to view
and manipulate XYZ point coordinates.
Finally, export the model to CAD,
animation, or rendering program or
continue adding more photos, points,
lines, and edges with PhotoModeler's
advanced marking tools to model NURBS
curves, cylinders, and surfaces.
 Camera System : White Light Scanning

Pair of
cameras

LCD projector that

throws strip patterns
on the object Video: Rivage - Beginning to End.avi
Video: White Light Scanning - ATOS1.avi
 Camera System : White Light Scanning
A real-time scanner using digital fringe projection and phase-shifting
technique (certain kinds of structured light methods) was developed, to
capture, reconstruct, and render high-density details of dynamically
deformable objects (such as facial expressions) at 40 frames per second

Different patterns can be applied to this system, and the frame rate for capturing
and data processing achieves 120 frames per second.

Typical accuracy figures are:

1. Planarity of a 2-foot (0.61 m) wide surface, to 10 micrometres (0.00039 in).

2. Shape of a motor combustion chamber to 2 micrometres (7.9105 in)
(elevation), yielding a volume accuracy 10 times better than with
volumetric dosing.
3. Shape of an object 2 inches (51 mm) large, to about 1 micrometre
(3.9105 in)
4. Radius of a blade edge of e.g. 10 micrometres (0.00039 in), to 0.4 m
 Camera System : White Light Scanning

20 March 2017 By Anil Gupta www.whitelightscanning.com

Camera Systems : Handiscan
 Digitizing Methods

2D/
Contact/ Destructive/
Method Slice/
Non-contact Non-destructive
3D
Manual measurement 2D/ 3D Contact Non-destructive
Profile projector 2D Non-contact Non-destructive
Touch probe mounted on a CMM 3D Contact Non-destructive
Laser scanning 3D Non-contact Non-destructive
Industrial Computer Tomography (CT) Slice Non-contact Non-destructive
Magnetic Resonance Imaging (MRI) Slice Non-contact Non-destructive
Ultrasonic scanning Slice Non-contact Non-destructive
Photogrammetry 3D Non-contact Non-destructive
White light scanning 3D Non-contact Non-destructive
Face milling and measuring Slice Both Destructive
 Comparison of Contact & Non-contact
Types
Contact Type Non-contact Type
+ Accurate - Less accurate
- Slow + Very fast
- Labor-intensive
- Cannot reach deep interior features
 Capture devices

CMM

Machine tool

Dedicated scanning
machine (Eg. Renishaws
Cyclone) 21
 Capture devices

Machine tool
Advantages :-
Machine may already exist
Fast when integrated
Familiar to user
Large scanning area
Disadvantages :-
Machine not cutting when scanning
Slow with touch-trigger probe
Retro of old machine often is difficult.
22
 Capture devices

CMM

Advantages :-
As it is meant for this
application, it is optimal.
Large scanning area
Disadvantages :-
CNC Machine tool
released for cutting
Slow data capture
Very accurate Few CMMs have laser
scanning capability

23
 Capture devices

Scanning machine
Advantages :-
Fast data capture
Small styli - fine detail
Leaves machine tool for cutting
Purpose built for scanning
Disadvantages :-
Limited working range
Limited to data digitising
24
 Cyclone
Scanning machine

Stand alone scanning system.

Supports contact & non-contact scanning
Performance 1000 - 3000 mm/min
Up to 1000 points per second prior to data
filtering. Video: Cyclone.avi
25
Coordinate Measuring Machine (CMM)
Huge Gantry-Type CMM
Make: DEA
The floor itself is used as the
table.
The probe or the laser scanner is
mounted on a gantry.
Accuracy of 0.0002 over an
envelope of 150 x 100 x 100.
 FARO Arm Portable Scanner
Normally CMMs have only 3 axes. But Faro arm will have 6 or 7 rotary
joints or axes. In other words, its kinematics is same as that of an
articulated or jointed-arm robot. Each joint has an encoder which helps
the controller keep track of its tips position.

FARO Arm recovers data on detailed features within its reach (up to 20
feet).

It finds application where

Equipment is too large or awkward to get accurate measurements.
Complex surfaces or profiles are difficult to measure with hand tools.
Relation between different features of equipment is difficult to measure.
Inspection with hand tools is prohibitively time consuming.

It is possible to move the Faro arm over the surface while the controller
records the positions.

It can be interfaced with a Laser Tracker or Theodolite system to

accurately recover details on much larger pieces of equipment.
FARO Arm Portable CMM
 Layered Milling and Scanning

Integrated milling and

scanning machine

Image
processing

Preparation of
soft replica
 Industrial CT Scan
Human bodies can be scanned using Computer Tomography (CT).
The result is a picture of the human body part in slices.

Computer Tomography (CT) uses radiation in the form of a highly

collimated X-ray fan beam to slice the object. This results in points
arranged in parallel planar slices or loops. CT was initially
developed for medical applications. Of late, it is finding industrial
applications.

When carrying out image processing of these slices, rings of points

can be obtained by marking the interfaces between the various body
parts. These rings of points are then combined to form CoPs.

Standard CT scanners achieve a resolution of 512 x 512 elements

within a layer.

Today, CT scan is slow due to the rotation of the emitter-receiver

assembly and hence gives only static images.
 Principle of CT Scan

X-Ray can penetrate most objects;

Each absorb it based on their material characteristics and X-Ray wave length.
The detector records the remaining intensity. Note that this is just a single value
and gives no clue about the objects along the path of the ray.
Principle of CT Scan

The X-Ray transmitter and

detector assembly is rotated
around the object making the
measurements of intensities
at regular angular intervals.

Using a very complex

transformation, P(,n) values
can be converted into the
loops.
Principle of CT Scan

The data obtained from all

the slices can be rendered in
the form of voxels.

This can further be

processed to obtain B-Rep
models.
Industrial CT Scan Example
An industrial computed tomography (CT) scan
conducted on an aluminum casting to identify
internal failures such as voids. All color
coordinated particles within casting are
voids/porosity/air pockets, which can
additionally be measured and are color
coordinated according to size.

Image-based finite element methods

Image-based finite element method (IBFEM)
converts the 3D image data from X-ray
computed tomography directly into meshes for
FEM. Benefits of this method include
modelling complex geometries (e.g. composite
materials) or accurately modelling 'as
manufactured' components at the micro-scale
Industrial CT Scan Example

3D reconstruction of a
disposable pepper grinder. Glass
in blue.
 Magnetic Resonance Imaging
Magnetic Resonance Imaging (MRI) is yet another technique used
in the medical field to produce high quality images of the inside of
the human body.

MRI images are obtained based on different tissue characteristics

by varying the number and sequence of pulsed radio frequency
fields in order to take advantage of magnetic relaxation properties of
the tissues. MRI differs from CT in at least two key aspects:
i. MRI measures the density of a specific nucleus
ii. The MRI measurement system is volumetric, i.e., interrogation of the
entire body, within the measurement volume, is done all at one time.

CT and MRI represent the finest resolution capability available in

diagnostic systems achieving volumetric resolutions. During the
scanning process, the patient is stepped through the measurement
plane 2-3 mm at a time. The information from each plane can be put
together to provide a volumetric image of the structure as well as
the size and location of anatomical structures. The scanned model
becomes a virtual volume of the patients bone(s).
 Principle of MRI
MRI is based on the principles of Nuclear Magnetic Resonance
(NMR), a spectroscopic technique used by scientists to obtain
microscopic chemical and physical information about molecules.
The technique was called magnetic resonance imaging rather than
Nuclear Magnetic Resonance Imaging (NMRI) because of the
negative connotations associated with the word nuclear in the late
1970's.
 Principle of MRI
The principle of operation of CT scan and MRI are very similar.

The patients is placed in a very strong but biologically harmless

magnetic field, and a microwave radio signal is beamed through
the body at a specific frequency which interacts with the water
molecules in the body, causing them to absorb a tiny amount of
energy. These molecules then produce their own little radio signals
as they re-emit the energy they absorbed from the excitation beam.

Computer software calculates the density of the body tissues at

each point in the slice and constructs an image. As with the CT
scanner, the software can manipulate the images to differentiate
tissue types as well as produce three-dimensional views,
longitudinal slices, and so forth. MRI, since it is basically imaging
water densities, is best for soft tissue and least useful for bony
structures (for which CT scanning is best).
 Principle of MRI

Magnetic Resonance Imaging (MRI) is yet another technique used in the medical
field to produce high quality images of the inside of the human body.
 Principle of MRI
X-Ray : Captures only bones
Harmful
3D imaging not possible

MRI : Captures soft objects such as tissues nicely

Safer. So used in children.
3D imaging possible

CT scan : Captures all objects with good clarity.

Safer than X-Ray but more harmful that MRI.
3D imaging possible

Ultrasonic : Captures soft objects but dynamically.

3D imaging possible
Not harmful at all
3D Reconstruction
 3D Reconstruction
The points obtained from the device could be
Simply a cloud of points (E.g.: photogrammetry)
Organized sets of points (E.g.: Points along a scan line can be treated
as a curve in cases such as optical scans. Similarly, the points are
available as loops and slices in CT, MRI and ultrasonic scans.)

Adequate planning of the scanning process and partitioning the scanned

boundary is extremely important as
handling the cloud of points is very tough. Any existing pattern among
them simplifies the process.
If a certain zone is of a known geometry such as fillet or cone, the
reconstruction will be easier, faster and more accurate.
It will be very frustrating at the time of reconstruction to realize that
some zones were not scanned.
In optical methods, the measurement of surface points close to its
tangent will not be accurate. Proper planning will help avoid this.

Reconstruction involves arranging the scanned points in suitable

topological fashion. The hierarchy is vertices, edges, faces and solids.
 3D Reconstruction
When the measurements are made from various direction, each resulting
patch needs to be oriented. This requires the knowledge of the view point
w.r.t. a global reference.
 Conclusions
Direct or contact methods are more accurate but laborious and indirect or
non-contact methods are faster but less accurate.

Some digitizing methods are suitable for any object. E.g.: CT scan, MRI
scan, Ultrasonic scan etc.

Some others, such as optical techniques, require different settings to

absorb the entire data. This requires
overlapping of surfaces during scanning
registering or assembling the various scans using appropriate
transformations.

Some interior hidden features cannot be captured at all using these

methods.

The knowledge of the scanning pattern helps a lot during reconstructions.

This knowledge helps in segmenting the cloud of points into more
organized arrays of points.