DETECTION AND DISTINCTION OF COLOURS USING COLOUR

SORTING ROBOTIC ARM IN A PICK AND PLACE MECHANISM

TRINESH T M

4th sem,Mtech IAR VIJAYAVITHAL BONGALE

Professor & Head

Malnad College Of Engineering
Department Of Mechanical Engineering
Hassan 573202,Karnataka Malnad College Of Engineering
Hassan 573 202, Karnataka
[email protected]
[email protected]

Abstract: Color sorting Robot is one of temperature white SMD LED is used for
the useful, costless and fastest systems in this purpose.The sensor assembly is
Industrial applications to reduce manual housed in a circular black opaque plastic
working time and provides less human ring to avoid external light interferences.
mistake when manual system is As per standard RGB schemes supported
undertaken. The objective of this project is by most computers and operating systems
to design an efficient, microcontroller the 16 bit value is converted to 8 bit and
based system that pick up right color of the resultant number of color which
objects and put it down at right place to TC34725 can differentiate totals to 16.7
optimize the productivity, minimizing the million colors. A low cost approach had
cost of the products and decreasing human been used to sense the size of the sample
mistakes. In our system the TCS34725 is object, wherein the requirements are that
interfaced with the Raspberry Pi A+ board the object should be opaque. The circuit is
on its 2nd I2C interface channel, running realized using a set of comparators – 8 x
Linux OS. The values are read using LM324 for each axis, and 2 x CMOS
python script by calling the python I2C counter type CD4017 to generate scanning
libraries. in both X and Y axis.
The circuitry is controlled by an 8051 core
The color sensor requires a light source to MCU (AT89S52). And is interfaced by its
illuminate the sample and a 5500K color UART to the Raspberry Pi A+ processor
board. The communication is handled in respective pre-programmed place.Thereby
high level by the pySerial libraries. eliminating the monotonous work done by
human, achieving accuracy and speed in
I.INTRODUCTION the work. The project involves colour
Determining real time and highly accurate sensors that senses the object’s colour and
characteristics of small objects in a fast sends the signal to the microcontroller.
flowing stream would open new directions The microcontroller sends signal to circuit
for industrial sorting processes. The which drives the various motors of the
present paper relates to an apparatus and robotic arm to grip the object and place it
method for classify in and sorting small- in the specified location. Based upon the
sized objects, using electronic systems and colour detected, the robotic arm moves to
advanced sensors operating on the basis of the specified location, releases the object
a physical and geometric characterization and comes back to the original position.
of each element. Recent advances in
electronics and printed circuit board II. LITERATURE SURVEY
technology open new perspectives for To reduce human efforts on mechanical
industrial application in this field.The maneuvering different types of robotic
proposed system is an embedded system arms are being developed. These arms are
which will increase the speed of color too costly and complex due to the
sorting procedure, provide the accurate complexity and the fabrication process.
color sorting process, decrease the cost of Most of the robotic arms are designed to
color sorting process and optimize the handle repeated jobs. In design of the
productivity of an industrial object. The robotic are different parameters are to be
system comprises of color sensor, stepper taken care. The design of mechanical
& servo motors and microcontroller. By structure with enough strength, optimum
another way this project can be treated an weight, load bearing capacity, speed of
automated material handling system & can movement and kinematics are important
be designed by following way. It parameters. In electronic design the
synchronizes the movement of robotic arm specification of the motors, drives, sensors,
to pick the objects moving on a conveyor control elements are to be considered.In
belt. the software side the re configurability,
It aims in classifying the coloured objects user interface and implementation and
which are coming on the conveyor by compatibility are to be considered. In
picking and placing the objects in its simple term, the reference sources
emphasize on few aspects like sorting of fixed start point where the sample object is
different colored objects can be done by placed. Depending on sorting criteria –
using camera, but here in this project deals Color, Size, Color & Size, the object can
with sorting of both different colored be picked and placed by the robotic arm at
objects and different size objects with the different location or different bins placed
help of advanced color sensor in a circular path around the robotic arm.
TCS34725FN.
B.Color Sensing
III..PROPOSED SYSTEM
BLOCK DIAGRAM OF THE SYSTEM The color sensing is carried out by a
semiconductor type reflective color sensor
manufactured by Austria Micro Systems –
TCS34725FN, the device features a 4
channel color sensor with built in Infrared
filter to cut off unwanted infrared
radiations pick up from surrounding light
sources. The device consists of an internal
A-D convertor for light to digital value
conversion and each channel can be read
using its I2C interface bus as a 16bit
digital value. These values are read by the
host micro-computer or processor to
analyze the RGB values and represent
them or process as per requirement. In our
system the TCS34725 is interfaced with
the Raspberry Pi A+ board on its 2nd I2C
interface channel, running Linux OS. The
values are read using python script by
calling the python I2C libraries.
A. Working Principle of the whole The color sensor requires a light source to
System illuminate the sample and a 5500K color
The system sorts different size and colored temperature white SMD LED is used for
objects by making measurements on a this purpose. The sensor assembly is
housed in a circular black opaque plastic
ring to avoid external light interferences. wherein the requirements are that the
As per standard RGB schemes supported object should be opaque. An array of 8x8
by most computers and operating systems IR phototransistors and LED’s are used to
the 16 bit value is converted to 8 bit and scan for an object in X and Y axis and
the resultant number of color which evaluate the size in steps of 3-3.5mm, as
TC34725 can differentiate totals to 16.7 each component’s diameter is 3mm. The
million colors.As with any instrumentation circuit is realized using a set of
device a range calibration is always comparators – 8 x LM324 for each axis,
required and the same for this color sensor and 2 x CMOS counter type CD4017 to
is done by making measurements of generate scanning in both X and Y axis.
known color patches from shade cards, and The LED on one axis is scanned in a
substituting an error value computed by stepped manner and the return signal is
inverse matrix multiplication of three monitored at the photo transistor end,
different elements Red, Green, and Blue which is converted to a logic level signal
measured by the sensor, thus any other by the comparator. Thus any blockage of
color measured will be multiplied with an the IR light falling from source to target
error matrix to get compensated color represents the X and Y axis area occupied
values reproducing real color of the sample in multiples of 3mm by the sample. The
scanned. circuitry is controlled by an 8051 core
The error mainly occurs due to light source MCU (AT89S52). And is interfaced by its
used for illuminating the sample, the UART to the Raspberry Pi A+ processor
spectral distribution of the light source and board. The communication is handled in
its color temperature, combined with the high level by the pySerial libraries.
IR cut off filter present in the color sensor
lowers the color temperature and thus the
scanned colors reflect a blue tinge,
however after error correction the sample
color and scanned color can be rated with a
D.Pick and Place Control
delta E value of 1.5 to 2.0 which is very
The robot is controlled using 3 x servo and
close to the real color.
1 x stepper motor for complete pick and
place operation. The servo motors take
C.Size (Dimension) Sensing
PWM pulses from the AT89S52 MCU for
A low cost approach had been used to
varying and maintaining their position, the
sense the size of the sample object,
stepper motor is controlled by a IR spectral component of the incoming
transistorized circuit based on a TIP 127 light and allows color measurements to be
PNP darlington pair power transistor. made accurately. The high sensitivity,
The stepper motor is used to rotate the arm wide dynamic range, and IR blocking filter
to required angle in 3.75 degree step. An make the TCS3472 an ideal color sensor
optical slotted switch is used to sense solution for use under varying lighting
home position and end position to stop the conditions and through attenuating
motor movement at both ends. materials. This data is transferred via an
The complete mechanical assembly is cut I2C to the host
out of an acrylic 6mm panel with laser FEATURES
cutting process and same is joined using  Integrated IR blocking filter
fastening screws and cyanoacrylate  3.8M:1 dynamic range
adhesive. The specific material is used  Four independent analog-to-digital
based on its light weight and high tensile converters
strength.  A reference-channel for Color
Analysis (Clear channel photo-diode)
BENEFITS
IV Description of Hardware  Minimizes IR and UV spectral
Components and Software component effects to produce accurate
This part of the paper is showing the main color measurement
idea of the working principle of each
 Enables accurate color and ambient
components connected in the circuit to
light sensing under varying lighting
achieve the needed aim and objectives.
conditions
 Minimizes motion / transient errors
 Clear-Channel provides a reference
allows for isolation of color content

A.COLOR SENSOR TCS34725FN

B. COMPARATOR LM324
The TCS3472 device provides a digital
The LM324 series consists of four
return of red, green, blue (RGB), and clear
independent, high gain,internally
light sensing values. An IR blocking filter,
frequency compensated operational
integrated on-chip and localized to the
amplifiers which were designed
color sensing photodiodes, minimizes the
specifically to operate from a single power • Input common- mode voltage range
supply over a wide range of voltages. includes ground
Operation from split power supplies is also • Differential input voltage range equal to
possible and the low power supply current the power supply voltage
drain is independent of the magnitude of • Large output voltage swing 0V to V+ -
the powersupply voltage. 1.5V
Application areas include transducer
amplifiers, DC gain blocks and all the C. MICROCONTROLLER 8051
conventional op amp circuits which now (AT89S52)
can be more easily implemented in single The AT89S52 is a low-power, high-
power supply systems. For example, the performance CMOS 8-bit microcontroller
LM324 series can be directly operated off with 8K bytes of in-system programmable
of the standard + 5V power supply voltage Flash memory. The device is manufactured
which is used in digital systems and will using Atmel’s high-density nonvolatile
easily provide the required interface memory technology and is compatible
electronics without requiring the additional with the indus-try-standard 80C51
– 15V power supplies. instruction set and pinout. The on-chip
Flash allows the program memory to be
FEATURES reprogrammed in-system or by a
• Internally frequency compensated for conventional nonvolatile memory pro-
unity gain grammer. By combining a versatile 8-bit
• Large DC voltage gain 100 dB CPU with in-system programmable Flash
• Wide bandwidth ( unity gain) 1 MHz on a monolithic chip, the Atmel AT89S52
(temperature compensated) is a powerful microcontroller which
• Wide power supply range: Single supply provides a highly-flexible and cost-
3V to 32V or dual supplies – 1.5V to – effective solution to many embedded
16V control applications.
• Very low supply current drain ( 700 á A) The AT89S52 provides the following
— essentially independent of supply standard features: 8K bytes of Flash, 256
voltage bytes of RAM, 32 I/O lines, Watchdog
• Low input biasing current 45 nA timer, two data pointers, three 16-bit
(temperature compensated) timer/counters, a six-vector two-level
• Low input offset voltage 2 mV and offset interrupt architecture, a full duplex serial
current: 5 nA port, on-chip oscillator, and clock
circuitry. In addition, the AT89S52 is maximum speed, or to link Python
designed with static logic for operation programs to libraries that may only be
down to zero frequency and supports two available in binary form (such as a vendor-
software selectable power saving modes. specific graphics library). Once you are
The Idle Mode stops the CPU while really hooked, you can link the Python
allowing the RAM, timer/counters, serial interpreter into an application written in C
port, and interrupt system to continue and use it as an extension or command
functioning. language for that application.

D. IR PHOTOTRANSISTOR V. APPLICATIONS
The ITR9908 consist of an infrared The system has many applications in
emitting diode and an NPN silicon various fields, as this system provides the
phototransistor, encased side-by-side on sorting of objects, in flow of objects by
converging optical axis in a black multi sensing. Mainly this finds the
thermoplastic housing. The phototransistor important application in agriculture field
receives radiation from the IR only . This where it can be used to sort the different
is the normal situation. But when an agriculture products like grains, lemons,
reflecting object close to ITR , almonds, grapes, and different kind of
phototransistor receives the reflecting fruits. For human beings it becomes
radiation . comber sum task to sort out the objects
with high quality also the possibility of
E.SOFTWARE REQUIRED PYTHON: accuracy is less. In industry it can be used
Python is a general-purpose interpreted, for sorting of various objects and tools
interactive, object-oriented and high-level with different sizes. By this way the
programming language. Python was proposed project can be used. It finds
created by Guido van Rossum in the application in enormous way in
lateeighties and early nineties. Like Perl, agriculture, industry.
Python source code isalso now available
VI.Conclusion
under the GNU General Public License
(GPL).
The present work when implemented
Python is extensible: if you know how to
provide good re-search knowledge on
program in C it is easy to add a new built-
robot arm modeling and embedded based
in function or module to the interpreter,
control hardware and software
either to perform critical operations at
implementation provides an easier access Synchronized Robotic Arm”, International
to exercise robot manipulation using the Journal of Research in Engineering and
functionalities and programming abilities Technology, Volume: 02 Issue: 08, Aug-
of the real robots for mounting different 2013.
industrial applications. The objects with [4] Yeow Khang Yung, ‘’Color Sorting
different color can be determined by using System with Robot Arm’’ Faculty of
advanced color sensor TCS34725FN and Electronic and Computer Engineering
objects with different size also can be University Technical Malaysia Melaka,
determined by arrary matrix .Finally all the (2011)
objects are picked and placed by pick and
place robot with efficient manner. It [5] Malvin Nkomo, Member of SAIEE and
reduces the time and cost of investment is Michael Collier ‘’A Color-Sorting
also very low as compared to other robots. SCARA Robotic Arm’’ College of
Information and Electrical Engineering
Shandong University of Science and
References
Technology Qingdao, (2010), China
[1] Aji Joy, “Object Sorting Robotic Arm
Based on Colour Sensing ’’, Assistant
Professor, Department of Electronics and
Communication Engineering, Mar
Athanasius College of
Engineering, Kothamangalam, Kerala,
India, Vol. 3, Issue 3, March 2014

[2] C. Chandra Mouli and K. Nagabhushan

Raju ‘’A Review on Wireless Embedded
System for Vision Guided Robot Arm for
Object Sorting’’ , International Journal of
Scientific & Engineering Research,
Volume 4, Issue 7, 33 ISSN 2229-5518,
July-2013.

[3] GoldyKatal, Saahil Gupta,

ShitijKakkar,”Design And Operation Of