ABSTRACT
In this dissertation we developed a fiber optic connector optical return loss (ORL) model from the connector endface image profile of a 200X inspection microscope. The connector endface profiles were histogram equalized and then evaluated with the Canny, Sobel, Prewitt, Laplacian of the Gaussian (LoG), Roberts and Frei-Chen edge detection operators using Matlab® 2013 script, these operators yielded edge pixel count averages of 139.1, 49.43, 49.57, 103.2, 45.19 and 50.98 detections per section respectively. The Canny operator emerged the best candidate technique after yielding the maximum edge pixel count of 139.1 detections per section, where the figure of merit is the magnitude of the total number of edges detected. The image segmentation and image similarity measure of the Image under Test (IUT) and the Measurement Quality Jumper (MQJ) were implemented using the cross correlation coefficient and pixel based segmentation method respectively so as to establish the coefficients of the connector profile region reflectance in the ORL model. These tools were implemented using a combination of the Matlab 2013 Graphic User interface (GUI), image processing toolbox and symbolic math toolbox. The model results showed a mean connector ORL accuracy of ±0.5dB and ±0.55dB for the multimode and Singlemode connector ORL respectively. These fall within the TIA/EIA-455-95-A maximum specification of ±1dB and ±5dB for the Singlemode and Multimode optical fibers.
TABLE OF CONTENTS
Declaration ………………………………………………………………………………………………………….i
Certification ………………………………………………………………………………………………………..ii
Dedication ………………………………………………………………………………………………………….iii
Acknowledgement ………………………………………………………………………………………………iv
Table of Contents ………………………………………………………………………………………………..vi
List of Figures……………………………………………………………………………………………………..ix
List of Tables……………………………………………………………………………………………………….x
List of Abbreviations……………………………………………………………………………………………xi
Abstract………………………………………………………………………………………………………………xii
1.0 CHAPTER ONE: INTRODUCTION
1.1 Background. ……………………………………………………………………………………………1
1.1.1 Introduction to Optical Fiber Communications System. …………………………………2
1.1.2 Digital Image Operations. …………………………………………………………………………..3
1.2 Aim and Objectives ………………………………………………………………………………….5
1.3 Statement of Problem ………………………………………………………………………………5
1.4 Methodology ……………………………………………………………………………………………6
1.5 Dissertation Outline …………………………………………………………………………………8
CHAPTER TWO: LITERATURE REVIEW
2.1 Introduction …………………………………………………………………………………………….9
2.2 Review of Fundamental Concepts …………………………………………………………….9
2.2.1 Wave Propagation. …………………………………………………………………………………….9
2.2.2 Optical Fiber profiles and modes. ……………………………………………………………….10
2.2.3 Optical fiber attenuation ……………………………………………………………………………11
2.2.4 Optical Return Loss (ORL) ……………………………………………………………………….12
2.2.5 Edge Detection ………………………………………………………………………………………..19
2.2.6 Image Convolution theorem ………………………………………………………………………22
2.2.7 Image Histogram ………………………………………………………………………………………28
2.2.8 Image Segmentation …………………………………………………………………………………29
2.2.9 Image Similarity Assessment ……………………………………………………………………..33
2.3 Review of Similar Works ………………………………………………………………………….41
3.0 CHAPTER THREE: MATERIALS AND METHODS
3.1 Introduction ……………………………………………………………………………………………46
3.2 Image Histogram……………………………………………………………………………………..46
3.3 Edge Detection Operators………………………………………………………………………..47
3.4 Pixel-Based Segmentation. ………………………………………………………………………49
3.5 Development of Connector Endface Return loss model equation. ……………..49
vii
4.0 CHAPTER FOUR: RESULTS AND DISCUSSION
4.1 Introduction………………………………………………………………………………………………..51
4.2 Edge Detection Methods Evaluation…………………………………………………………….51
4.3 Fiber optic Connector ORL Model ……………………………………………………………..53
4.4 Connector ORL Measurement GUI. …………………………………………………………..56
4.5 Validation of Connector ORL Model……………………………………………………………57
5.0 CHAPTER FIVE: CONCLUSION AND RECOMMENDATIONS
5.1 Conclusion………………………………………………………………………………………………….60
5.2 Significant Contributions…………………………………………………………………………….60
5.3 Limitation…………………………………………………………………………………………………..61
5.4 Recommendation for further work………………………………………………………………61
REFERENCES…………………………………………………………………………………………………….62
APPENDICES………………………………………………………………………………………………………68
Appendix A…………………………………………………………………………………………………………….68
Source Code for the Implementation Of Edge Detection Operators……………………………….68
Appendix B…………………………………………………………………………………………………………….70
Matlab Source Code for the Plot of Detected Edges per Section of Connector
EndfacProfile ………………………………………………………………………………………………………..70
Appendix C…………………………………………………………………………………………………………….71
Screenshot of the data statistics tool of plot of edge detections per section
of the connector endface profile. ……………………………………………………………………………….71
Appendix D……………………………………………………………………………………………………………72
Laplacian of Gaussian Filter Code ……………………………………………………………………………72
Appendix E…………………………………………………………………………………………………………….73
Image Profile Sections of Canny Operator Output……………………………………………………….73
Appendix F…………………………………………………………………………………………………………….76
Reference MQJ and IUT connector profile images and their OTDR measured
Connectors Return Loss…………………………………………………………………………………………..76
Appendix G……………………………………………………………………………………………………………78
MalabM-file source code for connector return loss equation parameters…………………………78
Appendix H……………………………………………………………………………………………………………79
Matlab Graphic User Interface [GUI] Source Code……………………………………………………..80
Appendix I……………………………………………………………………………………………………………..85
Singlemode Optical Return Loss Measurement GUI Interface. …………………………………….85
Appendix J……………………………………………………………………………………………………………..86
Multimode Optical Return Loss Measurement GUI Interface. ……………………………………..86
Appendix K…………………………………………………………………………………………………………….87
Setup of OTDR and Inspection microscope measurements. ……………………………………………….87
Appendix L ……………………………………………………………………………………………………………88
Typical Optical Fiber Cable Manufacturer‟s Specifications ……………………………………………….88
CHAPTER ONE
INTRODUCTION
1.1 Background.
Communication can be said to be one of the activities that is as old as human existence itself.
It has metamorphosed incredibly over the course of documented human history; from spoken
words, sounds, hand gestures, smoke signals etc. All through this evolution, the
communication model has remained largely the same with modifications made to each
functional block as shown in Figure 1.1 to leverage on prevailing technological advances.
INFORMATION
(Message)
INFORMATION
(Message)
RECEIVER
(Decoder)
CHANNEL
(Medium)
TRANSMITTER
(Encoder)
Figure 1.1: Generic Communication Network Model (James, 2004)
The use of light for telecommunication dates back to 1790 when a French engineer named
Claude Chappe designed one of the most practical applications of light in communication
called the optical telegraph (James, 2004). The optical telegraph was used to relay signals
over a distance of 230km in about fifteen (15) minutes. However the optical telegraph was
soon replaced by the copper telegraph. Researchers, however, remained undaunted in their
pursuit of developing an optical communication system and this resolve later paid off with
the development of the LASER (light amplification by stimulated emission of radiation) by
an American physicist named Theodore Maiman which earned him the appellation of
“Father of Electro-optical Industry”(Amandeep et al., 2013). There has been a steady growth
in the global adoption of fiber optics due to its large capacity, immunity to electromagnetic
interference, attenuation etc (Casimer, 2002).
2
Nigeria as a country has also not been left out in the adoption of fiber optic technology
following investments in the South Atlantic 3/West Africa Submarine Cable (SAT3/WASC)
in Lagos by the Nigerian Telecommunication Ltd (NITEL) and installation of a host of
backbone fibre rings in the cities of Lagos, Abuja and Kaduna etc.(Vivien, 2011). The
SAT3/WASC 120Gbits/second undersea cable provides a communications pathway between
Asia and Europe (Philip, 2011). Other privately owned undersea/ terrestrial fiber optic
cables have since been deployed such as the GLO-1, Main ONE and the terrestrial 10
Gigabit Campus Fiber Network of the Ahmadu Bello University, Zaria.
1.1.1 Introduction to Optical Fiber Communications System.
Optical fiber communications involves the propagation of modulated light along a fiber optic
wave guide based on the principle of total internal reflection. The modulation takes place at
the transmitter segment as shown in Figure 1.2 where the electrical signal is converted to
light pulses by either Light Emitting Diodes (LED) or Light Amplification by Stimulated
Emission of Radiation (LASERS) and demodulated at the receiver by photo detectors.
Electrical
Signal (in)
Connectors
Transmitter Receiver
Electrical
Optical Signal (Out)
fiber Link
Figure 1.2: A Typical Fiber Optic Link (Bill & Emile, 2005).
3
As the light pulses propagate, they are susceptible to back reflections or return loss. Laser-diode transmitters offer higher performance than LEDs, but they are also more sensitive than LEDs to light reflected back into them from the fiber optic communication system. The reflected light can change the wavelength of the transmitting laser and add noise to the transmitted signal (destructive interference). Reflections that enter a Vertical Cavity Surface Emitting Laser (VCSEL) affect lasing action in the cavity and add noise to the optical signal (Bill & Emile, 2005). The added noise introduces interference to the fiber optic link, thus leading to data loss and reduced information throughput (Berdinskikh et al., 2002).
1.1.2 Digital Image Operations.
An image is a 2-D representation of a three-dimensional scene. This 2-D representation is implemented by a two-dimensional function f(x, y) that represents a measure of some characteristics such as brightness or colour of viewed scene. This connotes that for each position (x, y) in the projection plane, f(x, y) defines the light intensity at that point (Jayaraman et al., 2011)
The manipulation of this 2-D function can be divided into three categories (Ian, 1998):
i. Image Processing: image in image out
ii. Image Analysis: image in measurements out
iii. Image Understanding: image in high-level description out
Image processing refers to those operations that can be applied to digital images to transform an input image a [m, n] into an output image b [m, n]. Where “m” and “n” are
integer coordinates with {m=0,1,2,…,M–1}columns and {n=0,1,2,…,N–1} rows, the intersection of a row and column is termed a pixel. These operations are classified into three categories namely (Ian, 1998):
4
Point Operation: This is an operation in which the value b[m, n] at a specific coordinate in
the output image is dependent only on the input value at that same corresponding coordinate
a[m, n] in the input image as illustrated in Figure 1.3.
Figure 1.3: Illustration of Point Operation.(Ian, 1998)
Local Operation: An operation is local when the output value at a specific coordinate b[m,
n] is dependent on the input values in the neighborhood of that same coordinate a[m, n] in the
input image as illustrated in Figure 1.4.
Figure 1.4: Illustration of Local Operation (Ian, 1998).
Global operation: A global image processing operation is one in which the output value at a
specific coordinate b[m, n] is dependent on all the values in the input image as illustrated in
Figure 1.5.
Figure 1.5: Illustration of Global Operation (Ian, 1998).
Image analysis on the other hand investigates the image data to gain insight into the
happenings within the image and apply this knowledge to specific applications. The process
requires the use of tools such as (Qiang and Roberts, 2015);
5
i. Image segmentation
ii. Image transformation
iii. Feature extraction and pattern classification.
Segmentation of gray scale images into regions for measurement or recognition is the most important single problem area for image analysis, however a vast array of techniques have been developed in implementing image transformation, feature extraction and pattern classification (John, 2011).
1.2 Aim and Objectives
This research aims at developing a grahic user interface (GUI) based method of measuring optical fiber connector Optical Return Loss (ORL) by exploiting the connector image profile of an inspection microscope. The objectives of the research are as follows:
i. Using the Matlab image processing toolbox, evaluate the performance of edge detection methods in identifying the edges on a connector end-face image profile.
ii. Deduce the mathematical relationship between the edges detected and the connector optical return loss and develop a Matlab graphical user interface [GUI] based application using the established relationship to measure connector return loss.
iii. Validate the results obtained using the measurement quality jumper (MQJ) as a reference.
1.3 Statement of Problem
It has been established from literature that multimode and singlemode fiber optic cables are susceptible to Rayleigh scattering (accounting for about 90% of total attenuation) which occurs when light collides with individual atoms in the glass knocking it off its original course (Bagad, 2009) while absorption (which accounts for between 3%-5% of total
6
attenuation) is caused by light being absorbed by residual materials such as metals or water ions within the fiber core and inner cladding (Vivek, 2004). The Rayleigh scattering and absorption induced attenuation are manufacturer dependent (Vivek, 2004). Return loss falls in the category of the remaining 5%-7% contribution to overall attenuation.
This dissertation seeks to extend the functionality of a typical fiber inspection microscope to not only give a visual display of the endface of fiber optic connector but also compute the connector optical return loss (ORL) which will enable the engineer to make objective assessment of the suitability of a field terminated optic fiber cable.
This is achieved by identifying and implementing using Matlab GUI, image processing toolbox and symbolic math toolbox the appropriate edge detection method, image segmentation and similarity measure to determine the connector optical return loss (ORL) from the connector endface image profile of a fiber inspection microscope thereby improving upon its contemporary function of visual inspection alone.
1.4 Methodology
The procedures implemented for this dissertation are:
i. Termination of three (3) sets each of both multimode and singlemode optical fiber cables using ST connectors.
ii. Using the 200X fiber optic inspection microscopes, insert the terminated fiber optic cable in step (i) into the microscope and capture the end-face image of the connector.
iii. Using an OTDR capture the reflective event trace and record the optical return loss for the multimode and singlemode fiber optic connectors respectively.
7
iv. Using the Matlab image processing toolbox and the captured end-face image profile from step (ii) transform these profiles with the variants of the following edge detection methods:
a. Canny method
b. Sobel method
c. Prewitt method
d. Roberts method
e. Laplacian of Gaussian method
f. Frei-Chen method
v. Plot and evaluate the output of these transforms so as to select the method that yields the most edges as the candidate method.
vi. Transform the connector end-face image profiles with the candidate method in step (v) and establish the relationship between the transformed image profile and the optical return loss measured in step (iii) above using the Matlab image processing toolbox commands.
vii. Develop a MATLAB Graphic User Interface application that uses the established relationship in step (vi) to perform measurements on new connector end-face samples.
viii. Validate the results from the algorithm to be developed by comparing it with that of the measurement quality jumpers (MQJ)
8
1.5 Dissertation Outline
The general introduction has been presented in chapter one. The other chapters are outlined as follows; the review of fundamental concepts of image processing and analysis closely followed by the review of similar research works was presented in chapter two. The edge detection variants were implemented and evaluated using MATLAB 2013 software in chapter three. In Chapter four the results of the edge detection operators and validation of the developed connector ORL model is presented and discussed. The conclusions drawn, limitations encountered, significant contributions and recommendations for further work were covered in chapter five.
IF YOU CAN'T FIND YOUR TOPIC, CLICK HERE TO HIRE A WRITER»
