ROAD WORK ZONE AUTOMATED TWO FLAGGERS SIGNALING SYSTEM IN LANE CLOSURE USING HAAR CASCADE ALGORITHM
A Thesis Presented to the Faculty of Information and Communications Technology Program STI College Southwoods
In Partial Fulfilment of the Requirements for the Degree Bachelor of Science in Computer Engineering
Carl Justine S. Canada Eugene A. Labrador Mark Andrey U. Leal Shairmaine Mae M. Managbanag
October 14, 2018
ENDORSEMENT FORM FOR ORAL DEFENSE
TITLE OF RESEARCH:
Road Work Zone Automated Two Flaggers Signaling System in Lane Closure using Haar Cascade Algorithm
NAME OF PROPONENTS:
Carl Justine S. Canada Eugene A. Labrador Mark Andrey U. Leal Shairmaine Mae M. Managbanag
In Partial Fulfilment of the Requirements for the degree Bachelor of Science in Computer Engineering has been examined and is recommended for Oral Defense.
ENDORSED BY:
Engr. James B. Labrada Thesis Adviser
APPROVED FOR PROPOSAL DEFENSE:
Engr. James B. Labrada Thesis Coordinator
NOTED BY:
Program Head
October 14, 2018
STI College Southwoods
ii
APPROVAL SHEET This thesis titled: Road Work Zone Automated Two Flaggers Signaling System in Lane Closure using Haar Cascade Algorithm prepared and submitted by Carl Justine S. Canada; Eugene A. Labrador; Mark Andrey U. Leal; and Shairmaine Mae M. Managbanag, in partial fulfilment of the requirements for the degree of Bachelor of Science in Computer Engineering, has been examined and is recommended for acceptance an approval. Engr. James B. Labrada Thesis Adviser
Accepted and approved by the Thesis Review Panel in partial fulfilment of the requirements for the degree of Bachelor of Science in Computer Engineering
<Panelists' Given Name MI. Family Name> <Panelists' Given Name MI. Family Name> Panel Member Panel Member
<Panelists' Given Name MI. Family Name> Lead Panelist
Noted:
Engr. James B. Labrada Thesis Coordinator
Program Head
October 14, 2018
STI College Southwoods
iii
ACKNOWLEDGEMENTS The developers would like to thank the following: We sincerely thanks to our Thesis Coordinator/ Thesis Adviser, Engr. James B. Labrada, for the continuous support, motivation and ceaseless knowledge to finish this thesis. For our Thesis Review Panel, for sharing expertise, encouragement, insightful comment and suggestions, We would like to express our gratitude to our Parents and/ or Guardian, for their support and sympathetic ears when in times of necessities in our thesis and when we have problems. And for our friends and inspirations, we are thankful for sharing ideas not only support but also happily by talking about things other than just our papers. Finally, special thanks also from our God above when in times of giving up and having problems.
STI College Southwoods
iv
ABSTRACT Title of research:
Road Work Zone Automated Two Flaggers in Lane Closures using Haar Cascade Algorithm
Researchers:
Carl Justine S. Canada Eugene A. Labrador Mark Andrey U. Leal Shairmaine Mae M. Managbanag
Degree:
Bachelor of Science in Computer Engineering
Date of Completion: <Month year of graduation> Key words:
The abstract is an important component of your thesis. Presented at the beginning of the thesis, it is likely the first substantive description of your work read by an external examiner. You should view it as an opportunity to set accurate expectations. The abstract is a summary of the whole thesis. It presents all the major elements of your work in a highly condensed form. An abstract is not merely an introduction in the sense of a preface, preamble, or advance organizer that prepares the reader for the thesis. In addition to that function, it must be capable of substituting for the whole thesis when there is insufficient time and space for the full text. Currently, the maximum sizes for abstracts submitted are 150 words to 350 words. Usually a one-pager abstract is the most ideal. To preserve visual coherence, you may wish to limit the abstract for your research to one 1.5-spaced page, about 280 words. The structure of the abstract should mirror the structure of the whole thesis, and should represent all its major elements. For example, if your thesis has five chapters (introduction, literature review, methodology, results, conclusion), there should be one or more sentences assigned to summarize each chapter. In the succeeding paragraphs, there should be no indentations, paragraphs are justified with left alignment. Delete this highlighted section and replace it with your Abstract.
STI College Southwoods
v
TABLE OF CONTENTS Page Title Page
i
Endorsement Form for Proposal Defense
ii
Approval Sheet
iii
Acknowledgements
iv
Abstract
v
Table of Contents
Vi
List of Tables
Vii
List of Figures
Viii
List of Appendices
Ix
1. Introduction
1
1.1 Background of the problem
2
1.1.1 General Problem
3
1.1.2 Specific Problem
4
1.2 Overview of the current state of technology
5
1.2.1 Flagman
6
1.2.2 Warning Signs
6
1.2.3 Channelizing Device
7
1.2.4 Arrow Display
7
1.3 Objectives of the study
8
1.3.1 General Objectives
8
1.3.2 Specific Objectives
8
1.4 Scope and limitations of the study
9
1.4.1 Scope
10
1.4.2 Limitations
21
2. Literature Review
22
2.1 Foreign Studies
32
2.2 Local Studies
27
2.3 Synthesis
29
3. Overview of the Project
STI College Southwoods
30
vi
3.1 System Design Specifications
32
3.2 Hardware
32
3.2.1Raspberry Pi 3 Model B
32
3.2.2 Raspberry Pi Camera v2
32
3.2.3 Signal Light
33
3.2.4 Buck Converter
33
3.2.5 Battery 12V
34
3.2.6 Solar Panel Charge Controller
35
3.2.7 Solar Panel
35
3.2.8 Through Beam Sensor
36
3.2.9 Radio Frequency Module
37
3.3 Software
39
3.3.1 OpenCV
39
3.3.2 Python
39
3.4 System Overview
41
3.4.1 System Hardware Design
41
3.5 Block Diagram
42
3.6 Flowchart
43
3.7 Cost
45
Results and Discussions Conclusions and Recommendations References Appendices
50
Appendices A
51
Appendices B
54
Appendices C
56
Appendices D
58
STI College Southwoods
vii
List of Tables Table 1
Page Summary of Foreign AFAD Implementation
STI College Southwoods
16
viii
LIST OF FIGURES Figure
Page
1
Sample of positive and negative image
2
2
Samples of Flagman
6
3
Samples of Warning Sign
6
4
Channelizing Device
7
5
Arrow Displays
7
6
Feature Calculation
15
7
Table Calculation of Images
16
8
Rectangle Features
16
9
Features
18
10
Sample of Humidity
22
11
Examples of AFADs
24
12
DRAAKSHA Flagman Robot
25
13
Advance Warning Signs
26
14
Sample of Self-driving Car using Advanced Computer Vision with OpenCV
27
15
Standard Two-Flagger Operations
28
16
Channelizing Devices- Barricades
29
17
Evolutionary Prototyping Model
30
18
Raspberry Pi
32
19
Raspberry Pi Camera 8MP
33
20
Signal Light
33
21
Buck Converter
34
22
Battery
34
23
Solar Panel Charge Controller
35
24
Solar Panel
35
25
Thru Beam Sensor
36
26
Circuit Diagram of Through Beam Sensor
36
27
Radio Frequency Module
37
28
RF Module to Raspberry Pi Connection
38
29
OpenCV Logo
39
STI College Southwoods
ix
30
Python Logo36
40
31
Design of the Prototype
41
32
Flagger Operation
42
STI College Southwoods
x
List of Appendices Appendix
Page A. List of revisions during proposal defense and demonstration
50
B. User’s Manual
54
C. User’s Acceptance Training
56
D. Curriculum Vitae of Researchers
58
STI College Southwoods
xi
INTRODUCTION Road construction is an essential process of development for the comfort and safety of users. Most road construction is a distillate job whereby the duration and safety are always considered for the construction management. One of the most crucial factors in a road construction site is the safety of the road users and the construction workers. Proper safety sign and indication system should be utilized to ensure safety to be optimized. The most common practice of safety indication is performed by traffic controller or also known as the flagman. The risk of fatal accidents is high for the flagman since he is placed in the highest risk zones. In recent years, driving safety has become a more important issue. Number of car accidents and casualties increase year by year. Based on developers’ research, the main reason of occurring accidents is human negligence. According to Philippines' Department of Public Works and Highway- Traffic Recording and Analysis System, 727 accidents in construction work zones and maintenance/ utility work zones. Therefore, this system could provide more information about the vehicle approaching to the driver, so that the driver could make the correct decision when driving on the road. The system could monitor whether the vehicles appear in the other lane or not and inform the driver when the car is near at the road work zone area. And it helps the contractors and workers doing their work without worries. Computer vision teaches computers to see, discovering, analyze and identify the existence of objects from surrounding environment using object detection algorithms. Haar Cascade Algorithm is an effective method for vehicle detection by collection of more positive and negative samples which will be used in the system (see Figure 1.1). A
true positive image is indeed present in the image processed and the classifier labels it indicates a positive result. A false positive falsely determines that the object is located in the image, even though it is not. A true negative image does not detect an object even it is present in the detection windows. A false negative is unable to determine the actual object from the image.
Figure 1. Sample of positive and negative image . Background of the problem The Department of Public Works and Highways (DPWH) maintains the national roads and bridges and aims to improve the maintenance of roadway by giving more emphasis on maintenance activities in order to preserve and keep the infrastructure as nearly as possible in its original condition. In this way, the government would get a good benefit from the huge investment on the roadway infrastructure and the drivers would be satisfied. It is of utmost importance that roadways are safe and always kept in a good condition for the protection of travelling public, the people in the vicinity, the road itself and all the road users in general.
STI College Southwoods
2
According to Department of Public Works and Highways (DPWH), most of the time incident happens in flagman’s due to reckless driving of road users. Some drivers enter the bypass lane at high speed than posted speed limits is one of the cause of loss control and results in road accident that will drift the construction workers and other road user into fatal zone. Flagman cannot stay for 24 hours to assist a vehicle in a road construction. At times the flagman’s concentration on controlling the traffic flow also can result in various miseries not only to his life but also to the public. The common cause of delay in a road construction is due to weather conditions and where illegal or unauthorized properties are in a road construction. So, the developers proposed a device which will ensure the security of the workers in 24 hours during road closures. The basic system requires sensors to detect the presence of vehicles on road construction site, a control unit to process the information and to establish communication between the sensor unit and display unit.
1.1.1
General Problem The general problem of the study entitled “Road Work Zone Automated Two- Flaggers Signaling System in Lane Closure using Haar Cascade Algorithm” is to design, develop and implement a signalling system that can be warn the approaching vehicle before entering in a road construction or in a two-lane road that reduced to single lane road.
STI College Southwoods
3
1.1.2
Specific Problem The purpose of the study is to develop the following:
1. How to develop a system that will help the highway workers in a road construction? Manual operation can also have a human error that can usually be attributed to a failure to manage that results in injury, costly damage to vehicles, equipment, and premises. The system will be the helping tool for the road and safety management system for highway workers.
2. How to develop a system that will determine the volume of the vehicle that passes by using Haar Cascade Algorithm? Traffic managing in single- lane closure in a two- way road is easy to manage if the flow is light because it is easy for flagman to count and decide which lane will go first and which lane will stop. But, when flagman encountered heavy traffic in a road construction, deciding and counting may cause traffic congestion because it takes time for flagman to count a vehicle in a heavy traffic.
STI College Southwoods
4
3. How to develop a system that will be more useful in night- time and day- time closures using Raspberry Pi Camera and signal lights? In night- time closures, most of the time road users or drivers hardly see the flagman or road works while in day- time closures, flagman can be easily get tired or exhaust due to perpendicular sunrays and weather conditions that causes accidents.
4. How to develop a system that will supply the device using solar power system? The system is needed to have a long-time power source using solar power system, that independent on main electricity or AC power source.
Overview of the current state of the technology 1.2.1 Flagman A person who has charge of carries, or signals with a flag, especially in road construction zones. They also indicate to the road users enters a bypass lane to be caution on the pavement which may not be as smooth as the expressway.
STI College Southwoods
5
Figure 2. Samples of Flagman
1.2.2 Warning Signs Signs used in road work zone traffic control. Regulatory signs impose legal restrictions and may not be used without permission from the authority with jurisdiction over the roadway. Warning signs give notice of conditions along the roadway.
Figure 3. Samples of Warning Sign
STI College Southwoods
6
1.2.3 Channelizing Devices The function of channelizing devices is to warn road users near the roadway and to guide road users. Channelizing devices include cones, tubular markers, vertical panels, drums, barricades, and temporary raised islands. Channelizing devices provide for gradual vehicular traffic flow from one lane to another.
Figure 4. Channelizing Device
1.2.4 Arrow Displays Arrow panels are traffic control devices used for additional advance warning. They are generally used for lane closure, and slow moving maintenance activities. Use flashing arrow panels for all lane closures on highways where the posted speed limit equals or exceeds 45 mph.
Figure 5. Arrow Displays
STI College Southwoods
7
Objectives of the study 1.3.1 General Objective The general objective of the study entitled “Road Work Zone Automated Two Flaggers Signaling System in Lane Closures using Haar Cascade Algorithm” is to design, develop and implement a system in a road construction that will signify the drivers to slowdown and move to the other lane especially in a two-lane road that reduced to single road.
1.3.2 Specific Objectives The specific objectives of the studies are guided by the following:
To develop a system that will help the highway workers in a road construction. The system will help the highway workers in a road construction to lessen the work of flagman and to manage the traffic flow faster to avoid vehicle congestions.
To develop a system that will determine the volume of the vehicle that passes by using Haar Cascade Algorithm. The system will determine the volume of the vehicle by using Haar Cascade Algorithm that will collect the positive and negative samples of images to process the flow properly and based on how
STI College Southwoods
8
many cars have already passed and will help in deciding which lane will go first.
To develop a system that will be more useful in night-time and day- time closures using Raspberry Pi Camera and signal lights. Raspberry Pi Camera will be serve as the eye of the system that will determine the volume of the vehicles. There will be adjustments in exposures and white balance to see the approaching vehicle even at night. The signal light will indicate the drivers if the vehicle will stop or go.
To develop a system that will supply the device using solar power system. The system must be solar powered system to lessen the electricity consumption and for longer duration.
STI College Southwoods
9
Scope and limitations of the study 1.4.1 Scope The scope of this project are as follows:
The system will be used in publicly and privately funded road construction projects which require overnight or 24-hour traffic control regardless of weather conditions.
The system will organize the flow of traffic. The system will organize the traffic flow using camera that will capture images then process by raspberry that will be the output of the traffic light. The timer will count up to 60 seconds to indicate the drivers to enter in a one-way lane.
The system is portable or capable only in a two- lane closures.
The two automated flaggers will be able to communicate within 1 kilometre from each other.
The system will handle traffic based on vehicle count seen by the camera. The system will prioritized the maximum volume of cars. If station A has 10 cars and station B has 5 cars, the one half of 10 cars in station A will go first. But in consideration of time and to avoid traffic congestion in both stations, if station A has 10 cars and
STI College Southwoods
10
station B has 3 cars, the one half of 10 is 5, and if it is below one half of 10, the station B will go first. And vice versa.
1.4.1.1 Input Module The following module will be served as input of the system:
Raspberry Pi Camera v2 This module served as the eye of the system as it will be the one in charge to capture approaching vehicles either day or night. The range sufficient to reach about 130 to 160 degrees wide and can detect the car with the maximum of 100 meters. The sensor itself has a resolution of 8 megapixel. In terms of still images, the camera is capable of 2592 x 1944 pixel static images, and also supports 1080p30, 720p60 and 640x480p60/90 video. The raspberry pi camera’s function is to capture approaching vehicles and compare the gathered image into positive and negative images. Positive images- target object which is the vehicle. Negative images without target objects. For example, empty road. It is also depends on the registered positive images.
Thru Beam Sensor This module would count how many vehicles passed on the first device. The maximum vehicle that will only passed is 10 vehicles. If the thru beam on the first device reached the number of vehicles passed, it must be similar to the second device and that will appear
STI College Southwoods
11
in the signal light as “Go”. Thru beam sensor has a maximum range of 100 meters away from the device.
1.4.1.2 Output Module The following module will be served as output of the system:
Signal Lights There are three colours that convey as an indicator for road users. The signal light output will be based on the data gathered in lane which is more crowded. When the lane is more crowded than the other lane, its signal light output is green while the other traffic light output is red which is stop.
1.4.1.2 Control Module These are the following modules that controls certain parts of the system:
Raspberry Pi The Raspberry Pi board is the central module of the whole embedded image capturing and processing system. The board supports 802.11n wireless LAN (peak throughput of 150Mbps). The technical specification will be 1 gigabytes (GB) and the single board computer running at 1.2 gigahertz (GHz).
STI College Southwoods
12
Radio Frequency Module The Radio Frequency module will be the bridge of communication between the two Raspberry Pi. The air data rate is the modulated signalling rate the nRF24L01 uses when transmitting and receiving data. The air data rate can be 1Mbps or 2Mbps and makes it possible to transmit over 1 kilometre. A 2.4 GHz signals required line of sight, any trees or buildings in the way will block the signal.
1.4.1.4 Software These are the following software applications and algorithm:
Python The main programming language to work with OpenCV that give instructions to Raspberry pi. It managed the memory used by the program. A program written in a complied language such as C or C++.
OpenCV The developers used an open source computer vision and it is part of Python’s software library that helps the Raspberry Pi to identify the existence of approaching vehicle on both lanes. OpenCV’s application areas include: 2D and 3D feature toolkits, gesture
STI College Southwoods
13
recognition, and object identification. OpenCV is written in C++ that binds in Python, Java, MATLAB/OCTAVE.
Haar Cascade Algorithm An effective method for vehicle detection by collection of more positive and negative vehicle samples. A true positive image is indeed present in the image processed and the classifier labels it indicates a positive result. A false positive falsely determines that the vehicle is located in the image, even though it is not. A true negative image does not detect the vehicle even it is present in the detection windows. A false negative is unable to determine the actual object from the image. It is a machine learning based approach where a cascade function is trained from a lot of positive and negative images. It is then used to detect objects in other images. The algorithm needs a lot of positive images (images of vehicles) and negative images (images without vehicles) to train the classifier. Each feature is a single value obtained by subtracting sum of pixels under white rectangle from sum of pixels under black rectangle.
STI College Southwoods
14
Figure 6. Feature Calculation Now all possible sizes and locations of each kernel is used to calculate plenty of features. For each feature calculation, to find the sum of pixels under white and black rectangles using integral images. Integral image is summed-area table is a data structure and algorithm for quickly and efficiently generating the sum of values in a rectangular subset of a grid. To understand this look at image 1 and image 2.
Image 1 is the source table, Image 2 is the
summation table. Notice in Image 2 row 1, col 2 value 33 is sum of Image 1 row 1 (col 1 + col 2).
STI College Southwoods
15
Figure 7. Table calculation of Images
Rectangle features can be computer using an intermediate representation for the image. The integral image at location x,y contains the sum of the pixels above and to the left of x,y.
Figure 8. Rectangle Features
STI College Southwoods
16
The sum of the pixels within rectangle D can be computed with four array references. The Value of the integral image at location 1 is the sum of the pixels in rectangle A. The value at location 2 is A + B, at location 3 is A + C, and at location 4 is A + B + C + D. The sum within D can be computed as 4 + 1 – (2 + 3). The first feature selected seems to focus on the property that the region of the eyes is often darker than the region of the nose and cheeks. The second feature selected relies on the property that the eyes are darker than the bridge of the nose. But the same windows applying on cheeks or any other place is irrelevant. Where ii(x,y) is the integral image and i(x,y) is the original image.
(Where s(x,y) is the cumulative row sum, s(x,y - 1) = 0, and ii(x 1,y) = 0) the integral image can be computed in one pass over the original image. Using integral image any rectangular sum can be computed in four array references. Clearly the difference between two rectangular sums can be computed in eight references. Since the two rectangle features define above involve adjacent rectangular sums the can be computed in six array references.
STI College Southwoods
17
Figure 9. Features
The first and second features selected by Adaboost. The two features are shown in the top row and then overlayed on a typical training face in the bottom row. The first feature measures the difference in intensity between the region of the eyes and region across the uppers cheeks. The feature capitalizes on the observation that the eye region is often darker than the cheeks. The second feature compares the intensities in the eye regions to the intensity across the bridge of the nose. To select the best features out of 160000+ features using Adaboost which both selects the best features and trains the classifiers that use them. Problems in machine learning often suffer from the curse of dimensionality each sample may consist of a huge number of potential features (for instance, there can be 162,336 Haar features, as used by the Viola–Jones object detection framework, in a 24×24 pixel image window), and evaluating every feature can reduce not only the
STI College Southwoods
18
speed of classifier training and execution, but in fact reduce predictive power, per the Hughes Effect. AdaBoost training process selects only those features known to improve the predictive power of the model, reducing dimensionality and potentially improving execution time as irrelevant features need not be computed.
1.4.1.5 Supply Module These are the following storage components that will be serve as power supply to the system:
12V Battery The main power supply would be from a 12V battery with 100Ah capacity. It is use as the main source of the system and the limit initial current to 30A. Charge until battery voltage reaches 13.90 to 14.20 volts at 68’F (20’C). Hold battery across constant voltage source of 13.50 to 14.20 volts continuously. When held at this voltage, the battery will seek its own current level and maintain itself in a fully charged condition.
Solar Panel Charge Controller The Solar Panel Charge Controller can control the battery recharging automatically and make sure the battery work longer and safer.
STI College Southwoods
19
A charge controller, basically a current regulator to keep batteries from over charging. It regulates the voltage and current coming from the solar panel going to the battery.
Buck Converter DC 12V to 5V Step Down The Buck Converter used as a converter from 12V to 5V for the supply of Raspberry Pi. It is a DC to DC power converter which steps down voltage. It is a class of switch mode power supply (SMPS), typically it is containing at least two semiconductors (a diode and a transistor) And it consist of LM2596S step down power module
20W 12V Solar Panel The solar panel served as the charger of battery.
STI College Southwoods
20
1.4.2 Limitation The limitations of this project are as follows:
The system cannot identify what type of vehicle approaching but can identify if it is a vehicle. The system uses Haar Cascade Algorithm which is only programmed to identify if it is a vehicle or not.
The system cannot record the plate number of the vehicle. The systems objective is to manage the flow of one lane road but in terms of emergency or evidence it cannot record the plate numbers.
The system cannot generate a report. It will affect the transmission between the transmitter and receiver of the device.
The system cannot be used in rainy weathers. The humidity of the area can affect the transmission of data that can cause accidents. The device will be switched into manual operation if the area is in high level of humidity to avoid accidents. The average monthly relative humidity varies between 71 percent to 85 percent. Typhoons or super typhoon will affect the transmission of data too.
STI College Southwoods
21
Figure 9. Sample of Humidity
STI College Southwoods
22
LITERATURE REVIEW Theoretical/Technical background 2.1 Foreign Studies 2.1.1Automated Flagger Assistance Device (AFADs) Automated flagger assistance devices (AFADs) are mechanically operated temporary traffic control devices that function under the same operational principles as traditional flagging. AFADs are considered a safety enhancement because they minimize flaggers’ direct exposure to traffic by allowing them to control the flagging device from an area away from traffic, such as behind a guardrail.
Table 1. Summary of Foreign AFAD Implementation
Shaded area indicates no information identified/provided *According to unpublished presentation. *According to manufacturer’s website.
a.) Boerne
c.) Floresville
b.) Tilden
d.) Uvalde
Figure 6. Examples of AFADs
All of the maintenance offices used AFADs remotely controlled at both ends of the work zone. Typically, a single qualified flagger operated both AFADs unless the flagger could not adequately view both AFADs and approaching traffic in both directions. All of the maintenance offices position the AFAD on the shoulder with the gate arm extended across the travel lane.
STI College Southwoods
24
2.1.2 DRAAKSHA DRAAKSHA is a human- like robot that waves hand similar to a flagman. Just like their human counterparts, waves hands to divert the traffic from one lane to another on expressways/ highways in events of road maintenance from main lane to side lane or vice- versa. Apart from highways and expressways, DRAAKSHA can be used in parking spaces to indicate the directions.
Figure 7. DRAAKSHA Flagman robot
STI College Southwoods
25
2.1.3 Advance Warning Signs
Figure 8. Advance Warning Signs
2.1.4 Self-driving cars using Advanced computer vision with OpenCV, finding lane lines The lines on the road wills how were the lanes are, will act as constant reference for where to steer the vehicle. A self-driving car is to automatically detect lane line using an algorithm.
STI College Southwoods
26
Figure 10. Sample of Self-driving car using Advanced computer vision with OpenCV
2.2 Local Studies The traffic control zone is the distance between the first advance warning sign and the point beyond the work area where traffic is no longer affected. Below is a diagram showing the parts of a traffic control zone according to DPWH.
STI College Southwoods
27
Figure 8. Standard Two-Flagger Operations
STI College Southwoods
28
Figure 9. Channelizing devices - barricades
2.3 Synthesis The developers in their attentive and careful data and requirements gathering on jeepney driver’s experiences and DPWH Engineer have identified that their study is feasible in software and hardware field aspects to aim the prototype complete. The information gathered help the developers to figure out of how the process works in two-way with single- lane closures and what standard level of materials for steady used of the prototype.
STI College Southwoods
29
ROAD WORK ZONE AUTOMATED TWO FLAGGERS SIGNALING SYSTEM IN LANE CLOSURE USING HAAR CASCADE ALGORITHM Overview of the project Evolutionary Prototyping Model aims to develop a mature system through a series of prototype iterations. The prototype will undergo a series of refinements, and should eventually become the solution. It is used when developers are faced with undefined or rapidly changing requirements.
Figure 9. Evolutionary Prototyping Model
3.1.1 Evolutionary Prototyping Model Phases
1. Identification of the Basic Requirement This phase is a continuous process, it should be able to identify the basic things including the input and the output information needed for the system to work.
2. Creating the Prototype This step is developed with each one better than the one before it to ensure it success that includes the user interface.
3. Verification of Prototype This step will be done through survey and experimentation using participants and examine the prototype and provide feedbacks by customers including end-users to find out if the system is good or not.
4. Revised and Enhanced the Prototype When the prototype is seen as inadequate, the developers will make further iteration using the feedback for both specifications and the prototype can be improved.
STI College Southwoods
31
System Design Specification 3.2 Hardware 3.2.1. Raspberry Pi 3 Model B The Raspberry Pi is a series of small and barebones computers. As the specifications, the Raspberry Pi is a credit card- sized computer powered by the Broadcom BCM2835 system-on-a-chip (SoC). The Raspberry Pi board is the central module of the whole embedded image capturing and processing system. Its main parts include: main processing chip unit, memory, power supply, HDMI. This main processing chip connects a camera and display.
Figure 10. Raspberry Pi
3.2.2 Raspberry Pi Camera v2 The Raspberry Pi Camera Module v2 is a high quality 8 megapixel camera based around the Sony IMX219 image sensor - it allows to create high definition video and still photographs.
STI College Southwoods
32
Figure 11. Raspberry Pi Camera 8MP
3.2.3 Signal Light This device will tell the drivers start driving or keep driving and stop. The light from a traffic light sealed-beam lamp in these programmable visibility signals passes through a set of two glass lenses at the back of the signal.
Figure 12. Signal Light
3.2.4 Buck Converter It is a class of switched-mode power supply typically containing at least two semiconductors and at least one energy storage element, a capacitor, inductor, or the two in combination. Buck converters can be highly efficient (often higher than 90%), making them useful for tasks such as converting a computer's main STI College Southwoods
33
supply voltage down to lower voltages. It is DC to DC converter and has a 15 watts and 12V to 22V and 3A to 5A.
Figure 13. Buck Converter
3.2.5 Battery 12V A device containing an electric cell or a series of electric cells storing chemical energy that can be converted into electrical energy, usually in the form of direct current. A 12V-25Ah rechargeable sealed electric battery is a device consisting of one or more electrochemical cells with external connections provide to power Raspberry pi.
Figure 14. Battery
STI College Southwoods
34
3.2.6 Solar Panel Charge Controller Solar Charge Controller can automatically manage the work of Solar Panel and battery in solar system. Solar charger controller can control the battery recharging automatically and make sure the batteries work longer and safer.
Figure 15. Solar Panel Charge Controller 3.2.7 Solar Panel 50 W 12V Solar panel designed to absorb the sun's rays as a source of energy for generating electricity or heating. It is used to charge a 12 volt battery.
Figure 16. Solar Panel
STI College Southwoods
35
3.2.8 Through Beam Sensor Through-beam sensors are distinguished by a long range. The system consists of two separate components: a transmitter and a receiver. The light only travels one way (from the transmitter to the receiver).
Figure 17. Through Beam Sensor
Figure 18. Circuit Diagram of Through Beam Sensor
STI College Southwoods
36
3.2.9 Radio Frequency Module An RF module (Radio Frequency Module) is a small electronic device used to transmit and/or receive radio signals between two devices. In an embedded system it is often desirable to communicate wireless communication through radio frequency communication. The module used was nRF24L01, a single chip 2.4 GHz transceiver and has 1.9 to 3.6 voltage supply range.
Figure 18. Nrf24L01
STI College Southwoods
37
Figure 1. RF Module to Raspberry Pi Connection
STI College Southwoods
38
3.3 Software 3.3.1 OpenCV OpenCV stands for Open Source Computer Vision Library, is an open source computer vision and machine learning software library. OpenCV was built to provide a common infrastructure for computer vision applications and to accelerate the use of machine perception in the commercial products. It can used to detect and recognize faces, identify objects, classify human actions in videos, track camera movements, track moving objects, and many more.
Figure 19.. OpenCV Logo
3.2.2 Python Python is an interpreter, object- oriented programming language similar to PERL that has gained popularity because of its clear syntax and readability. Python is said to be relatively easy to learn and portable, meaning its statements can be interpreted in a number of operating systems, including UNIX- based
STI College Southwoods
39
systems, Mac OS, MS DOS, OS/2, and various versions of Microsoft Windows 98.
Figure 20.. Python Logo
STI College Southwoods
40
3.4 System Overview 3.4.1 System Hardware Design Automated Flaggers can handle single lane closures, or can be combined with other signal systems to service side roads and driveways. The system is permanently mounted on a four-wheel trailer for easy transport and quick application. Let the system control the traffic flow at each site entrance and traffic lights with digital image processing using. The system’s battery activated and replenished during the workday with solar panels (see Figure 3.10). During road work zone, the two-flaggers will be placed both sides. The first Thru Beam Sensor purpose is to count how many vehicles passed and will compare on the other Thru Beam Sensor, if both Thru Beam Sensor has the same data gathered it means all the vehicle has passed. So, the device will determine when it will let the other vehicle go or stop (see Figure 3.11).
Figure 21. Design of the Prototype
STI College Southwoods
41
Figure 22. Flagger Operation
3.5 Block Diagram
The system is powered by solar panel which collects energy from sunlight and charge the battery. Once the system is turn on, the Raspberry Pi module will determine the volume of cars in one-way lane using the camera and pass through the Thru Beam Sensor. The system will determine that the other lane is congested; the device will indicate the drivers to go or stop using the traffic light.
STI College Southwoods
42
3.6 Flowchart 3.6.1 Haar Cascade Classifier
If the image fails the first stage, discard it. There is no consideration for the remaining features on it. If it passes, apply the second stage of features and continue the process. The window which passes all stages is a vehicle.
3.6.2 OpenCV and Haar Cascade Classifier
STI College Southwoods
43
Images or the real time video is captured from the camera placed in road work zone. This video is converted into number of frames. Converting the images into greyscale for easy to identify the vehicle when OpenCV and Haar Cascade Classifier is loaded. Each frame is compared with the pre-defined features of the Haar Cascade Classifiers. When the features are matched, the vehicle is detected and a square is drawn around the vehicle.
STI College Southwoods
44
3.7 Budgetary Estimate Below is the current breakdown of materials for DESPRO2. Quantity 2 2
Description Raspberry Pi Raspberry Pi Camera 8mp
Price 2,400.00 1,800
1
Buck Converter DC 12V to 5V Step Down
300.00
2
Battery
1,200.00
2,400.00
1
1,800.00
1,800.00
2
Solar Panel Charger Battery Panel Regulator Controller Solar Panel 50W 12v
1,600.00
3,200.00
2
Radio Frequency Module
300.00
600.00
1
Memory Card 16gb
550.00
550.00
1
Servo Motor MG995 – 160
720.00
720.00
1
LED NL-10MM- PC-10
60.00
60.00
1
Plain Sheet
800.00
800.00
1
F Bar
230.00
230.00
1
ML4
60.00
60.00
1
Hacksaw
180.00
180.00
1
Angle Bar
420.00
420.00
1
Tubular ( ¾ x ¾ x 1.5)
280.00
280.00
1
Tubular ( 1 x 1 x 1.5)
300.00
300.00
1
Tubular ( 1 ¼ x 1 ¼ x 1.5)
420.00
420.00
5
Capacitor 25V, 200µF
35.00
175.00
4
LM7805
25.00
100.00
10
Resistor 1000 Ω
1.00
10.00
3
PCB Board 2x4
25.00
75.00
1
Ferric Chloride
35.00
35.00
Foods
387.00
Fares
50.00
Total Amount
STI College Southwoods
Amount 4,800.00 3,600.00
21,252.00
45
RESULTS AND DISCUSSIONS In this section, you may include theoretical proof, verification or evidence. This should give an overview of the results from the experiments. Sample raw data shall be included to explain how these are presented. The full set of data shall be included as an appendix. This shall also contain a discussion of the information derived from the results, with sample raw data to support each result. You may use tables or figures. The numbering of the table should be continuous. Say you have Table 1 in your introductory part, and your next table appears in your methodology part. The first table in the methodology part should be numbered as Table 2 and so on. This is the same for numbering figures in your manuscript. Below is the format for the title of your table and figure: Table #. Only the first letter of the title is capitalized Figure #. Only the first letter of the title is capitalized In the succeeding paragraphs, there should be no indentations, paragraphs are justified with left alignment. Delete this highlighted section and replace it with your own results and discussions.
CONCLUSIONS AND RECOMMENDATIONS The first part of this section is your summary followed by the conclusion/s and last part is/are your recommendation/s. This section summarizes the results based on the results and discussion chapter. If there are only three specific questions or objectives, there are only three results summarized in this section and no presentation of tables or figures. A good summary should be comprehensive. A summary must be concise. Your summary should considerably be shorter than the source. Avoid repetitive details. A summary must be coherent and independent. You are expected to maintain your own voice throughout the summary. Don’t simply quote other researcher’s works instead use your own words to express your understanding of what you have read. After all, your summary is based on your interpretation of the findings points or ideas. However, you should be careful not to create any misinterpretation or distortion by introducing comments or criticisms of your own Conclusions should unite with the findings and accomplishments of the study. If there are three summaries, there should also be three conclusions. Conclusions are arranged as it appears in the findings. Moreover, rejection and acceptance of hypothesis, if applicable are explained under conclusion. Only conclusions which are based definitely from the findings or results should be made. Mere opinion which have no basis of facts and findings have no place in the conclusions of the study Recommendations are based on the conclusions. It may include further research of the study and/ or enhancement of the developed system In the succeeding paragraphs, there should be no indentations, paragraphs are justified with left alignment. Delete this highlighted section and replace it with your conclusion/s
REFERENCES American Road & Transportation Builders Association (ARTBA). Retrieved from https://www.workzonesafety.org/training-resources/fhwa_wz_grant/atssa_afad/ Minnesota Department of Transportation, Guidelines for the Selection of Temporary Lane Control Systems in Work Zones, Two-Lane Two-Way roads Closed to One Through Lane. Retrieved from http://www.dot.state.mn.us/trafficeng/workzone/doc/GuidelinesLaneControlSelection-Draft.pdf Oregon Department of Transportation, Automated Flagger Assistance Device. Retrieved from http://www.oregon.gov/ODOT/HWY/CONSTRUCTION/QPL/Docs/automated_flagger.pd f?ga=t North Carolina Department of Transportation, Automated Flagger Assistance Devices (NorthCarolinaSpecialProvisions). Retrieved from http://www.ncdot.gov/doh/preconstruct/wztc/DesRes/English/SpecialProv/AFAD_SP.pdf Traffic and Parking Control Company, Incorporated, Automated Flagger Assistance Device. Retrieved from https://www.tapconet.com/store/product-detail/z3bb/automated-flaggerassistance-device-afad?c=Xnab&sku=131711 Safety Technologies, Auto Flagger 54. Retrieved from http://www.autoflagger.com/products/autoflagger-54/ Highway Supply, LLC. (2016). Retrieved from http://highwaysupply.net/about-us/ Coral Sales Company, Highway Safety Solutions, (2018). Retrieved from http://www.coralsales.com/products/its/workzonesafety/northamericantrafficafad/ Department of Public Works and Highways, Highway Safety Design Manual (2004). Retrieved from http://www.dpwh.gov.ph/dpwh/references/guidelines_manuals/highway_safety_design_standard s_manual Institute of Traffic Management. Retrieved from http://www.mmda.gov.ph/course-offering/39itm/1683-special-course-for-road-construction-workers-flagmen Flagger Training Manual, Department of Transportation. Retrieved from https://www.codot.gov/library/traffic/work-zone-safety-and-work-zone-trafficoperations/flagger-program/flagger-program-documents/flagger-manual Raspberry Pi Camera Module components.com/raspberrypi
Datasheet.
(2018).
Retrieved
from
https://www.re-
Nrf24L01 Single Chip 2.4GHz Transceiver Product Specification v2_0. Nordic Semiconductor. (2007). Retrieved from https://www.nordicsemi.no ArduCam Raspberry Pi Camera Module 1/4- inch 5-megapixel module datasheet (2015). Retrieved from https://www.ArduCAM.com Siemens signal heads traffic solutions. (2008). Retrieved from https://www.siemens.de/traffic Rapid Object Detection using a boosted cascade of simple features. Retrieved from . https://www.cs.cmu.edu/~efros/courses/LBMV07/Papers/viola-cvpr-01.pdf Deep Learning Haar Cascade. Retrieved from http://www.willberger.org/cascade-haarexplained/
STI College Southwoods
49
APPENDICES
APPENDIX A. LIST OF REVISIONS DURING PROPOSAL DEFENSE AND DEMONSTRATION
STI College Southwoods
51
Unit Testing 1. Raspberry Pi Camera
STI College Southwoods
52
2. Signal Light with Relay Module
STI College Southwoods
53
APPENDIX B. USER’S MANUAL
STI College Southwoods
54
Insert here the pdf copy of your User’s Manual. You may use one or more pages for your user’s manual. Ensure that all necessary information are included.
STI College Southwoods
55
APPENDIX C. USER ACCEPTANCE TRAINING
STI College Southwoods
56
Insert here the pdf copy of your User’s Acceptance Training. You may use one or more pages for your user’s manual. Ensure that all necessary information are included
STI College Southwoods
57
APPENDIX D. CURRICULUM VITAE OF RESEARCHERS
STI College Southwoods
58
Curriculum Vitae of
<email address> contact number either cellular phone or landline or both
Level Tertiary TechVoc High School Elementary
Inclusive Dates month year month year month year month year
EDUCATIONAL BACKGROUND Name of school/ Institution
PROFESSIONAL OR VOLUNTEER EXPERIENCE Nature of Experience/ Inclusive Dates Job Title month year month year month year month year
Name and Address of Company or Organization
Listed in reverse chronological order (most recent first).
AFFILIATIONS Inclusive Dates month year month year month year month year
Name of Organization
Position
Listed in reverse chronological order (most recent first).
SKILLS SKILLS
Level of Competency
Date Acquired month year month year month year
TRAININGS, SEMINARS OR WORKSHOP ATTENDED Inclusive Dates Title of Training, Seminar or Workshop month year month year month year month year Listed in reverse chronological order (most recent first).
STI College Southwoods
59