Chap 1

Robots Classification:
1. Drive Technology:
1. Depending upon the source of power to operate the joints of the robot, the drive systems can
be electric, Hydraulic and pneumatic.
1. Electric Drive System:
▪ Electric drive system uses DC servo motor or DC stepper motors
▪ It provides moderate torque and precise speed control.
▪ Today most of the robotic manipulator uses electric drive system.
2. Hydraulic Drive System:
▪ Hydraulic Drive System uses compressed fluids.
▪ It is used for high torque, high speed manipulation and to lift heavy loads.
▪ It does not have cleanliness.
▪ Application: molten steel handling and auto body part handling.

3. Pneumatic Drive System:

▪ Pneumatic Drive System uses compressed air to operate end-effector.
▪ They are light weight, high speed and precise control.
2. Work Envelope Geometry:
2. the work envelope of a robot is defined as the locus of all point in 3D workspace which can be
reached by the robotic end-effector.
3. Motion Control Methods:
3. motion control methods are:
1. Point to point motion:
▪ In this method, the end effector follows a sequence of discrete points in the
work space.
▪ the path between two points are given by the user.
▪ Example: spot wielding, pick and place, loading and unloading.
2. continuous or controlled path motion:
▪ in this method, the end effector follows a prescribed path along the fixed
tragedy in work space.
▪ the speed of motion along the path may vary.
▪ Example: arc wielding, spray painting, & gluing.
Robots Application:
1. Manufacturing industry – robots are used in manufacturing industry for repetitive task such
as packaging and material handling.

2. Food and beverage industry - robots are used in food and beverage industry for task such as a
packaging, palletizing and sorting.

3. Aerospace industry - robots are used in aerospace industry for task such as wielding, painting
and assembly.

4. Automotive industry - robots are used in Automotive production for task such as welding
painting and assembly.

5. Mining industry - robots are used in Mining industry for task such as drilling, excavation and
exploration.
Robots Characteristics:
1. autonomy - robots are design to perform a task without any human intervention.

2. speed - robots can perform task at a much faster rate as compared to humans which leads to
increase in productivity and reduces time consumption.

3. Precision- robots are equipped with sensors and control system which allows them to
perform a task with high precision.

4. Flexibility - robots can be programmed to perform a wide range of task and can easily be
reconfigure to perform a new task.

5. repetition - robots are designed to perform a task again and again without error

Robots Specifications:
1. No. of Axes - it is the number of independent rotational or translational movements that a
robot is capable of. Types: major axes (x, y, z) and minor axes (yaw, pitch, roll)

2. Reach – it is the maximum distance that the robotic end-effector can reach from its base.

3. Stroke - it is the maximum distance that a robotic end-effector can travel along each axes.

4. Tool orientation – it refers to the orientation end-effector which can be controlled using minor
axes and major axes.

5. Repeatability – it is the ability of the robot to position its end-effector at same place again and
again.

6. Precision – it is the 'fineness' in the output of the sensor.

7. Accuracy – it is the ability of the robot to position its end-effector at desired location with min.
error.

8. Operating environment - it is the environment within which the robot has to perform the
task. The robot should be designed to operate in extreme temperature & contaminated air.

9. Payload Capacity - it is the maximum weight that a robot can carry and manipulate.

10.Maximum speed – it is the highest speed that a robot can achieve while performing the task.
Number of Axes:
4. The number of axes is the number of independent rotational or translational movements that
a robot is capable of.

Major Axes:
● The major axes are used to establish position of robotic wrist.
● It corresponds to the translational movements of the robot.
● Major axes are:
o X-axis – it is the horizontal axis in which positive values increases to the right and
negative values increases to the left.
o Y-axis - it is the vertical axis in which positive values increases upward and negative
values increases downward
o Z-axis - it is the depth axis in which positive values increases towards the viewer and
negative values increases away from the viewer.

Minor Axes:
● Minor axes are used to establish orientation of the end-effector.
● It corresponds to the rotational movements of the robot.
● Minor axes are also known as degrees of freedom (DOF).
● Pitch, roll, and yaw are three rotational axes that are used to rotate the end-effector in
different directions.
o Pitch – it is rotation around the X-axis, which corresponds to tilting the end-effector in
up and down direction.
o Roll – it is rotation around the Y-axis, which corresponds to rolling the end-effector from
one side to another side.
o Yaw – it is rotation around the Z-axis, which corresponds to turning the end-effector in
left or right direction.

6 DOF:
5. 6 DOF robots consists of 3 major axes (x, y, z) and three minor axes (yaw, pitch, roll).
6. 6 DOF robots are the most common type of robots used in industry because they are capable
of performing a wide range of tasks with high accuracy.
7. There are also robots with more than 6 DOF which are used for specific applications were high
degree of dexterity is required.

Reach:
 8. Reach is the maximum distance that the robotic end-effector can travel from its base.
9. Types:
1. Horizontal Reach – it is the maximum radial distance that a robotic end-effector can
reach.
2. Vertical Reach - it is the maximum elevation above the work surface that a robotic
end-effector can reach.
10.Reach is always greater than or equal to stroke.

Stroke:
11.Stroke is the maximum distance that a robotic end-effector can travel along each axes.
12.Types:
1. Horizontal Stroke – it is the total radial distance that the wrist can travel.
2. Vertical Stroke – it is the total vertical distance that the wrist can travel.

Tool Orientation:
 ● Tool orientation refers to the orientation end-effector.

● It can be controlled using minor axes (pitch, roll, and yaw) and major axes (X, Y, and Z axis).

● Major Axis determines the shape of the work envelope.

● Minor axis determines the type of orientation of the tool.

● By adjusting these parameters, the robot can position its end-effector in specific orientation

● If three independent minor axes are available, then the arbitrary orientation in 3D workspace
can be obtained.

Repeatability:
 ● It is the ability of the robot to position its end-effector at same place again and
again.
● A robot might give repeatability error due to problems like backlash in the gears and
flexibility in the links.
● A robot which is repeatable might not be very accurate.

Precision:
● Precision is the 'fineness' in the output of the sensor.
● It is the measure of the spatial resolution with which the tool can be positioned within
the work envelope.
● Example: a sensor which reads 2.1178 is more precise than the sensor which
reads 2.1 for the same physical variable.

Accuracy:
● It is the ability of the robot to position its end-effector at desired location with minimum
error.

Chap 2
Direct Kinematics:
● Direct kinematics problem is a problem in robotics and computer animation which involves
finding the position of the robotic end-effector.

● It maps the joint angle of a robot to its corresponding Cartesian coordinate in work space.

● Direct kinematics problem can be solved analytically or mathematically, depending upon the
complexity of the robot and the desired accuracy.

● The solution is typically represented using transformation Matrix, trigonometry and various
mathematical techniques which describes the position and orientation of the end-effector
with respect to the base.

Kinematic Parameters:
 ● kinematic parameters are a set of physical parameters that are used describe the geometric
and motion-related characteristics of a robot.
● A robotic manipulator can be assumed as a chain of rigid links which are connected by either
revolute joints or prismatic joints.
Parameters are:

1. Joint Parameters:
● Joint parameters are physical properties and characteristics of a robotic joint.
● the position and orientation of a two consecutive links is specified by the joint
parameters.
● joint parameters are:
i. Joint angle (θk) - it is the angle between a two adjacent links.
ii. Joint distance (dk) - it is the distance between a two adjacent links.
● For revolute joint, Joint angle will be variable and Joint distance will be fixed.
● For prismatic joint, Joint angle will be fixed and Joint distance will be variable.

2. Link Parameters:
● link parameters are the physical properties and characteristics of robotics links which
connects two consecutive robotic joints.
● Link Parameter are:
i. Link Twist angle (αk) – it is the angle between two adjacent joint axes.
ii. Link Length (ak) – it is the physical length of each link.
● The link parameters are constant for both revolute and prismatic joints.
Direct Kinematics 2DOF:
Direct Kinematics 3DOF:
 Chap 3
Inverse Kinematics: [DIA PENDING]
● Inverse kinematic problem is a problem in robotics and computer and animation which inverse
finding the joint angles of the robot.

● This method helps to determine and move the robotic end-effector to desired position and
orientation.

● It is challenging to solve inverse kinematic problem especially for Complex Robotics system
which have multiple DOF.

● It can be solved using various mathematical methods such as numerical optimization and
matrix algebra.
Why inverse kinematics solution is never unique?
● Inverse kinematics solutions are not always unique.
● It means there may be multiple sets of joint angles that can result in the same end-effector
position and orientation

Reasons why inverse kinematics solution are not unique:

1. Redundancy: Many robotic systems have more degrees of freedom than necessary to achieve
a particular task. This redundancy can result in multiple solutions for a given end-effector
position and orientation.

2. Singularities: Singularities occur when the manipulator reaches to a configuration where it

loses one or more degrees of freedom, resulting in a breakdown of the inverse kinematics
solution. At singularities, there may be infinite solutions or no solution at all.

3. Geometric constraints: Certain geometries of the robot and its environment can impose
constraints on the possible configurations that can achieve a desired end-effector position and
orientation. These constraints can result in multiple solutions or no solution at all.

4. Numerical errors: The numerical methods used to solve the inverse kinematics problem can
introduce numerical errors, which can result into multiple solutions or no solution at all.

5. Elbow Position: the robotic manipulator can only be in elbow up or elbow down position,
depending on the quadrant in which the robot lies. 1st quadrant for elbow up, 4th quadrant
for elbow down
Inverse Kinematics 2DOF:
Inverse Kinematics 3 DOF:
 Chap 4
Workspace fixtures:
● Workspace fixtures are accessories of the robot.
● It forms a part of the robot to perform a particular task.
Work space fixtures used in the robot are:
1. Part feeders:
● Part feeder are used to feed a part or an object to the robot for manipulation like
griping/grasping/picking.
● In order to grasp the object, it must be presented to the robot in such a way which is
compatible with the tool.
● It is arranged in a concentric manner around the robot base.
● It can be activated electrically, electromechanically and hydraulically.

2. Conveyors:
● They are linear transport device.
● They can transport the object in either from left to right or right to left manner.
● They are used for multi robot work cell.
● Advantage: it can accommodate more no. of robots as compared to carousels.
● Disadvantage: part or object cannot be swapped between two robots.
3. Carousels:
● They are rotary transport device.
● They transport object for manipulation back and forth between two or more robots.
● Robots are mounted in a circular fashion around the rotary turn table.
● Advantage: part can be swapped between two robots.
● Disadvantage: it can accommodate only few robots.

4. Fixed Tool:
● Fixed Tool are used to hold a sub-part in exact position and orientation.
● Most of the robots are one-handed and have only one tool for operation. If any
complicated task needs to be done (like putting a nut onto a bolt), then fixed tools
are used to secure the sub part.
● It is fixed and cannot move by itself.
● Examples: jigs, clamps etc.
Continuous path motion: (CP)
● In continuous path motion, the path is specified by the user, after which the robot moves
continuously in CP motion.

● This motion is performed by continuous path robots.

● When the robot end-effector moves along the path, care has to be taken to see that each &
every joint angle is controlled. Otherwise, the robot tool will not move along the path.

● When the robot tool performs the tasks, at that time each & every axes moves at different
velocities which is controlled by a computer.

Interpolated motion:
● In interpolated path motion, the user specifies the start point, end point & the via points
instead of specifying the path.

● Then, a trajectory planning software is used to interpolate between the knot points & via
points to generate a smooth trajectory T passing through all knot points.

● It is done by using continuous path motion control technique.

Different methods of splitting a trajectory:

1. 4 - 3 - 4 Trajectory:
● In 1st segment, from pick point to lift off point is defined by 4th degree polynomial.
● In 2nd segment, from lift off point to set down point is defined by 3rd degree
polynomial.
● In 3rd segment, from set down point to place point is defined by the 4th degree
polynomial.
2. 3 – 5 - 3 Trajectory:
● In 1st segment, from pick point to lift off point is defined by 3rd degree polynomial.
● In 2nd segment, from lift off point to set down point is defined by 5th degree
polynomial.
● In 3rd segment, from set down point to place point is defined by the 3rd degree
polynomial.
3. Cubic Trajectory (3 – 3 – 3 - 3 - 3):
● In this, each segment is defined by a 3rd degree polynomial.

Interpolated motion with parabolic blends:

 ● In interpolated path motion, when the robot passes through the knot points, the velocity
changes abruptly.
● This results in a jerky robotic motion which could burn out the drive circuit or damage the
gear.
● In order to solve this problem, we introduce a parabolic curve or blend which smoothly
connects the adjacent linear segments, resulting into a smooth path.
● A parabolic blend consists of three segments: acceleration segment, deceleration segment &
constant velocity segment.
● In acceleration and deceleration segments, the velocity of the robot changes when it passes
through the parabolic curve.
● In constant velocity segment, the velocity of the robot remains constant when it passes
through the linear segments.

Path:
 ● path is a route which is taken by the robot to move from source (pick point) to destination
(place point).
● Equation for path w(q) = Represented by 3rd/4th/5th order.
● It is purely spacial information.
● It does not depend on velocity and time.
● It is used in path planning algorithms.
● Example - Path between two points on a map.

Trajectory:
● Trajectory is a path with time information.
● if we specify times at which the robotic end effector should be at various points along the
path, then the path is converted into Trajectory.
● It is both spatial and temporal information.
● It depends velocity and time.
● It is used in motion planning algorithms
● Example - Trajectory of a robotic arm during assembly task
 Path Trajectory
Path is a route which is taken by the robot to
move from source to destination. Trajectory is a path with time information

It does not depends velocity and time It depends velocity and time

It is used in path planning algorithms It is used in motion planning algorithms

It is spacial information It is both spatial and temporal information

Example - Path between two points on a map Example - Trajectory of a robot arm during
assembly task
 Cartesian co-ordinate trajectory Joint space trajectory planning
planning

1 It plans the trajectory of the end-effector in It plans the trajectory of the robotic joints in
Cartesian space (x, y, z). joint space.

2 It is complex to implement. It is simple to implement.

3 Accuracy is less. Accuracy is more.

4 It takes into account the existence of the It does not takes into account the existence of
obstacle. the obstacle.

5 Input - desired end-effector position and Input - desired joint angle values
orientation

6 Application - Pick-and-place operations Application - welding and painting

Pick and Place operation: (Four-point pick and place
trajectory)
● The main function of robotic manipulator is to pick up an object from a particular place and
place it at another particular place in desired position and orientation.
● The taken by the robot to pick and place an object is known as four-point pick and place
trajectory.

Pick and place operations:

1. Pick Point:
● the point from where the object has to be pick up is known as pick point.
● it is the 1st operation of pick and place operation.
● it represents the initial position and orientation of the object.
● The pick point is always the center of gravity of an object.
● Gross motion is used from initial position to pick point.
● matrix:

2. Lift off Point:

● it is the point from where the robot actually starts moving before picking up the object.
● it is the 2nd operation of pick and place operation.
● This point is just above pick point by distance u.
● Fine motion is used from pick point to lift off point.
● Matrix:

3. Set Down Point:

● it is the point from where the robot actually starts moving before placing the object on the
work surface.
● it is the 3rd operation of pick and place operation.
● This Point is just above place point by distance u.
● Gross motion is used from Lift off Point to Set Down Point.
 ● Matrix:

4. Place Point:
● the point at which the object has to be placed is known as place point
● it is the 4th operation of pick and place operation.
● it represents the final position and orientation of the object.
● Fine motion is used from Set Down Point to Place point.
● Matrix:

Chap 5
Grasp Motion Planning:
● Grasp Motion planning is the process of identifying and selecting suitable grasp
configuration for the robotic end-effector to manipulate the object.
● It is important because it determines whether the robot can successfully pick up and move the
object without causing any damage.
● 3 main considerations while selecting a suitable grasp configuration for an object:
1. Safe Grasp Configuration: (Safety)
● A safe grasp configuration means that the robot should not come in contact with
other parts of the payload.
● The object should not come in contact with any obstacle.
● The robot should be safe at initial and final grasp configuration.

2. Reachable Grasp Configuration: (Reachability)

● A reachable grasp configuration means that the robot should be able to reach the
initial grasp position and when the object is in the hand, the robot should be able
to find a collision – free path to the final grasp position.
● Grabbing the object from the top works fine, but while placing it in an inverted
fashion might create a problem because the robot collides with the work surface.
Therefore, this method is not feasible.
● To solve this problem, the tool can grasp the object from the sides, then perform a
roll operation of the tool in either clockwise or anticlockwise direction and then
place it on the work surface.

3. Secured or Stable Grasp Configuration: (Stability or Security)

● Stable or secured grasp configuration means that the object should not slip, move
or fall down during the part transfer or during the mating operations.
● When the object is in the hand of end-effector, the gripping force should be such
that the object should not fall, slip, rotate or slide from the fingers of the
end-effector.

Fine motion planning.

 ● Fine motion planning is the process of planning of precise movements of the robot to
achieve a desired task.

● Fine motion is the motion in which an object comes in physical contact with the surrounding
environment of the robot.

● Various forces act on the part during the fine motion planning which includes frictional force,
gravitational force, and the force with which it is moving.

● These forces can be controlled using advanced modelling and control techniques.

● Techniques for obtaining fine motion planning are:

1. Impedance control technique

2. Generalized damper approach

3. Slow movement during picking and placing the object.

4. Guarded motion

5. Compliant motion

Gross motion planning:

 ● Gross motion planning searches all the available path in the free workspace for the robot.
● The space between the obstacles are known as freeways through which the robot can move
freely.
● Translations are done along the freeways and rotations are done at the intersection of the
freeways.
● gross motion is planned using Generalized Voronoi diagram (GVD).

Generalized Voronoi Diagram (GVD):

● GVD is basically a map of the environment which shows the path through which the robot
can move freely without colliding any obstacle.
● GVD divides an environment into regions based on the distance between two or more
obstacles in the environment.
● there are three types of interaction of the part with obstacle:
o Interaction between a pair of edges: when two edges interact with each other, an angle
is made at the point of contact. The GVD path is the angle bisector of this point of
contact along which the tool tip should move.

o interaction between a pair of vertices: when two vertices interact with each other, the
GVD path is a straight line which passes between the two vertices which the tool tip
should move.
o interaction a between a vertex and an edge: when a vertex and an edge interact with
each other, the GVD path is parabolic which the tool tip should move.
Task planning simulation problem:
● When a robot has to do a particular operation, the task has to be planned and for that we
need various data like trajectory plan, types of obstacles, types feedback data and etc.

● These data are known as task specifications which are given as input to the task planner and
the output is a planned motion.

● The task planner performs various types of simulation.

Source & Goal Scene:

● The input data for the task planner has to be in the form of images.

● Two types of image data are given as input the task planner:

o Source scene: it specifies the initial layout of the parts.

o Goal scene: it specifies the final layout of the parts.

● The task planner uses the information from the source and goal scene and information from
the feedback sensors to rearrange the parts in source scene corresponding to the goal scene.

● Task planning operation is represented by:

Example:
TASK PLANNING SUB PROBLEMS:
● When a task needs to be performed by a robot it is known as task problem.
● When the task is complicated it is a divided into sub task or sub problems
These sub problems are:
1. Scene analysis:
● If the vision system cannot identify the parts in the Source scene, then it will create an
error in the goal scene.

2. Part ordering:
● Part ordering sub problem is the order in which parts needs to be moved in source scene in
order to achieve the desired goal state.

3. Gross motion planning:

● In gross motion planning, we have to plan the proper moves of the object such that it
does not collide with another object.

4. Grasp Planning:
● Grasp Planning sub problem is the problem associated with the handling of the object.

5. Task Planner Simulation:

● Various data and specifications are given as input to the task planner.
● The output of task planner is a sequence of instructions to the robot to perform the
specified task.
● While doing task planning simulation, the given task is broken down into sub tasks and
each task is taken into consideration

Task Level Programming:

 ● Programming of a robot means to teach it how to do a particular task efficiently.

Methods of Programming:
1. Powered Lead Through Method:
● In the powered -lead through mode, a small control unit called as a teach pendant is
used.
● Using the teach pendant, each axis of the robot is moved manually, until the
combination of all the axial positions comes the desired position.
● The operator uses a series of push buttons on a teach pendant to give commands.
● All these operations are stored in memory while teaching the robot to do a particular
task which is later used to perform the specified task.
● Disadvantages:
▪ It is time consuming as it requires a human operator to program the robot.
▪ Robot might go out of service during the teaching period.
▪ It cannot be used for continuous motions because it is difficult to use the teach
pendant to perform complex geometric motions.
▪ It cannot be used with external sensors such as camera to get adaptable with the
surrounding environment.
▪ objects that are manipulated should be presented to the robot at the same
position and orientation every time.

2. Manual Lead through Mode of Programming:

● In this method, the programmer physically grasps the robotic arm & the end – effector
and manually leads it through the desired path at the required speed, while
simultaneously recording the continuous position of each axis.
● The manual lead through method is generally used for the Continuous Path (CP) robots.
● It is also known as walk through method.

3. Textual Method of Programming:

● In this method, the programmer instructs the robot to do certain task in English like
commands.
● The programmer is not concerned about the motion of the individual joints, because it
is calculated by a language processor like a compiler or interpreter.
● Offline Programming:
▪ In an offline programming, the programmer writes the program for the robot and
then tests that program offline.
▪ Once the program is executed successfully offline, then the programmer tests the
program on the robot.
 ▪ Disadvantages of the Off-line programming:
o Offline programming requires 32 bit computers to simulate robot movements
graphically. It also requires a software package.
o Offline programming creates positional errors that are caused by the
deflection of the manipulator links.
o The programmer has to specify details like manipulation task, layout of the
parts, movement attribute such as speed, acceleration, force, etc.

4. Task Level Programming Languages:

● In this method, the programmer has to describe the task in a high -level language which
is known as the task specification.
● The task specification is given as the input to the task planner.
● The task planner then uses a database called the world model to characterize the robot
& its environment.
● The task specification is divided into a sequence of sub-tasks by the task decomposer.
The subtasks are then passed through the sub-planner which generates an output.
● The task planner produces an output as a detailed set of instructions which is called the
robot program.

Task Planner: (Robot Controller)

 ● The robot task planner is used to find a path from the initial robot world to the final robot
world using a sequence of primitive operations.
● It uses simulation software and requires a large amount of information to complete the task.

Working:
● Image analyzer –
o It processes the raw images & converts the raw image data into digital data.
o The digital data is given as input to the task planner.

● Task Planner –
o Specifications about the task is given as input to Task Planner.
o The output of the task planner is a robot program to complete the task.
o Once the motion is planned in the form of a discrete tool configuration trajectory [Wk],
it is sent as input to the trajectory planner.

● Trajectory planner –
o The trajectory planner converts the discrete tool configuration trajectory [Wk] into a
continuous time joint space reference trajectory r(t) using various interpolation
techniques and inverse kinematics equations.
o The joint space trajectory is given as input to the robot controller.

● Robot Controller -
o It uses a torque regulator to generate the required torque profile T(t) to implement the
guarded and compliant motion, thereby successfully completing the manipulation task.

Chap 6
Segmentation:
● The process in which items in an image are separated from the background and also from
each other is called segmentation.
● It this, the image is partitioned into connected regions.

Methods to achieve Segmentation:

1. Thresholding:
● Thresholding is a technique to convert a grayscale image into a binary image.
● In this technique, a threshold value is chosen.
● All pixel values above the threshold value are assigned as value 1 and while those below
the threshold value are assigned as value 0.
● The resulting binary image contains only two pixel values which is black and white,
● It becomes easy to identify and separate objects.

2. Region labelling:
● In Region labelling, after segmentation of the image, the regions are assigned with
labels.
● In this technique, each pixel is assigned a label based on its location and connectivity to
neighboring pixels.
● Pixels with the same label are part of the same region.
● Different regions are assigned with different labels.

Iterative Processing:

● Removal of noises from an image by using iterative operations is known as iterative

processing.
 ● Stray noise can occur due to various factors such as improper illumination & shadows.

Methods of Iterative Processing:

1. Shrink operators:
● Shrink operators are iterative operators which converts 1 into 0.
● It converts foreground pixel which is present at the background of an object into background
pixel.
● Example: consider an image of a circular object with some noise in the background. On
applying shrink operator on the image, the noise is removed and the size of the object is
reduced while preserving its circular shape.
● This is useful when the overall size and shape of an object is more important for further
analysis.

2. Swell Operators:
● Swell operators are iterative operators which converts 0 into 1.
● It converts background pixel which is present at the foreground of an object into foreground
pixel.
● Example: consider an image of a binary letter "A" with some gaps in its lines. On applying
swell operator on the image, the gaps are filled and the lines are made thicker, resulting in a
complete and recognizable letter "A".
● This is useful when small details and fine structure of an object are important for further
analysis.
Structured illumination:
 ● If proper illumination of the object is not done, then the image captured will contain a lot of
noise.
● A good lighting system can project 3D information about the object like height and surface
texture.

Types of Lighting Schemes are:

1. Back-Lighting:

● In Back lighting, the object is illuminated from the backside of the object.
● For back lighting, the object should be opaque.
● Back lighting produces good contrast between the object and the background.
● It does not give information about the height of the object.
● Back-Lighting can only be used for thin and flat objects.

2. Front lighting:

● In front lighting, the lights are used behind the camera.

● Front lighting is used when we want more information about object.
● Front lighting produces low contrast and therefore care must be taken while arranging lighting
of the background to produce maximum contrast.
3. Side Lighting:

● In Side Lighting, the lights are arranged between the camera and the object at a certain
angle and only on 1 side.
● side lighting is used when we want to illuminate the surface texture of the object.
● It is useful for detecting bumps or dimples.
Corner Point Detection:
● Corner point detection is the process of identifying the corners points of an object.

● Corner points are locations in an image where there is a high change in contrast.

● It is needed to identify the corners of an object in order to pick it up or manipulate it.

● It helps the robot to align its end-effector with the object's edges and pick it up without
causing any damage.

● For binary images a technique called corner-point encoding is used to extract the vertex pixels
directly.

Working:
● Each pixel has eight adjacent pixels which is called its neighbors.

Pixel P with its 8 corner points

● To determine whether a given pixel is a corner point, we examine the intensity pattern of its
neighboring pixels.

● This is done by scanning the image with a set of 3 x 3 corner-point templates.

● In corner-point template, 0 represents background pixels and 1 represents foreground pixels.

● Below shown are corner-point templates which can be used to find corner point.

● The above corner-point templates are generated by considering the corner pattern which
appears at the upper right portion of the first template.

● Then it is rotated in anti-clockwise direction by multiples of pi/4 to generate the remaining

seven templates.
FORMULA FOR SUMS: