Since we finished our exams at 28/5/2008, we started our preparations to Imagine Cup world finals. We traveled to alexandria during the last week to work on the project. We were confused in some parts of the project, such as the final step in our project which is the pattern recognition; we tried Back-Propagation Neural Networks and Support Vector machine. Support vector machine was very impressive that it's accuracy was 93% and in some times reached 100%, while Neural Networks were only 73.4%, so we chose Support vector machine of course. Now we are trying to integrate and package our final work in all the project.

During our idea research we found no real systems implementing the idea, all are talking about best algorithms and test sites; no real work. So, we found that our project provides many features that are not included in only one system, and those are:
1. Providing scanning ability in 2-D and 3-D GPR images.
2. No system until now tried to use Support Vector Machine while it provides a perfect accuracy. 3. We take the best algorithm from all researches we found and integrated them together, to form a very powerful system.

- We won the first place in Imagine Cup competition, and we are going to France isA. Does anyone can believe that. Really we are very happy and so excited.

Here is the links that contains the details:
http://www.elakhbar.org.eg/issues/17471/0112.html

http://www.masrawy.com/News/Technology/General/2008/april/14/egypt-microsoft.aspx

http://www.egynews.net/wps/portal/!ut/p/c0/04_SB8K8xLLM9MSSzPy8xBz9CP0os3hzMxMjkwAXIwMDbx8jAyNLQ8egYBNzo2ATE_2CbEdFANYT_Ag!/?WCM_GLOBAL_CONTEXT=/wps/wcm/connect/migration/ertu/news_science_technology/2008041411274496502277

http://www.shobiklobik.com/forum/topic.asp?TOPIC_ID=125939

http://www.link0777.com/Articles/TECHNOLOGY/default.aspx?url=http://news.link0777.com/News/technology/2008/april/14/egypt-microsoft.aspx

http://www.arabnet5.com/index.asp?c=2&id=13809

http://elbashayer.com/v1/news_detail.php?id=2722&cate_id=39

- We now are trying to focus in order to achieve our main target, which is passing the last term in our academic life (isA) successfully, then get prepared to represent our beloved country ever, EGYPT, in the finals. we are supposed to compete with other 59 team from 59 country all over the world.

- Our ambition is not limited only to imagine cup, but also we are seeking to win Made In Egypt competition.

- The 10th meeting

Attendees:
1. Marwa Moharram
2. Ahmad Galal
3. Ramy Essam
4. Mahmoud Hassan
5. Samer Refaat
6. Abd Allah Hazaa

Group Action Items:-
1. Requirement Specification document revision.

- The 9th meeting

Attendees:
1. Ahmad Helmy
2. Marwa Moharram
3. Ahmad Galal
4. Ramy Essam
5. Mahmoud Hassan
6. Samer Refaat
7. Abd Allah Hazaa

Group Action Items:-
1. Imagine Cup orientation.

2. Requirement Specificaion document rebuilding and issues.

- The 8th Meeting

Attendees:
1. Ahmad Helmy
2. Marwa Moharram
3. Ahmad Galal
4. Ramy Essam
5. Mahmoud Hassan
6.Samer Refaat

Group Action Items:-
1. Blog issues
a.Add Box.net widget to hold our work on th blog.

3. Effort estimation issues
a. Reformatting and rebuilding Effort Estimation document.

- Again we meet at the seventh meeting.

Attendees:
1. Ahmad Helmy
2. Marwa Moharram
3. Ahmad Galal
4. Ramy Essam
5. Mahmoud Hassan

Group Action Items:-
1. Blog issues
a. Remove our telephones from blog.
b. Remove documentation from blog and upload them into any server and put links to it.

2. Bug Tracking system issues
a. Add all members into it.
b. Write requirements as modules into “Feature request”.
c. Put blog issues as bugs in the bug tracking system.
d. Put MATLAB code into “CVS”.

3. Effort estimation issues
a. Fix hours’ calculations & equations.
b. Prototype is a use cases with screen shots.

4. Requirement specification issues: a. Security can be (Save program logs – User authentication).

5. Project Plan issues
a. Remove unneeded processes.
b. Specify project life cycle (may use Water Fall model).
c. Specify Deliverables (Project – Requirement – User manual).
d. Specify needed skills and available needed skills to know the needed skills.

6. Try to make a draft of the schedule/ time plan

- Sixth meeting was not too much longer, our supervisors collected the needed documents that was mentioned last meeting and were intending to send to us. we was advised to create a bug tracking system to handle our project modifications and specifying roles more clear. Another advise was about using what is called visual source safe (VSS); this tool is supposed to act as a server to keep our code safe and to avoid code merging meetings, and that is accomplished by allowing each team member to see the last modified code and give him the ability to update in it then put it again in the server, so that always keep the last updated code at the server.

- Fifth meeting was cool; that we had revised Effort Estimation, Requirement Specification Document (Update), and Project Plan Document. Our supervisor saw that we need another bunch of documents to make :(, she said that she will look for them; those other documents were Test Plan, Conceptual Design, Logical Design, and Physical Design.

- Here we back again with the fourth meeting, this meeting was all about revising Requirements Specification Document from the sponsor side. Our supervisor explained what we couldn't figure out with ourselves, that's to be Distribution List, Stakeholders, Mandating Verbs, Business Backgroud, Auditing, Compliance Requirements, Assumptions & Constraints, and Appendix. That is only what we couldn't figure out :)

- 3rd meeting: Our supervisors tend to explain some documents in this meeting such as requirement document, time plan, project plan, and effort estimate. At the end of the meeting we were assigned to accomplish some of the previous documents in order to deliver next Thursday (7/2/2008). Our assignment is to fill two documents: Requirement document and Effort Estimate document.

- 2nd meeting: This time our supervisors, Eng/Ahmed Helmy and Eng/Marwa Moharram, gave us a brief description about there working process and they told us that they will give us the company's templates to work on. They also told us that our work must be well documented in order to accomplish any project successfully.

- 1st meeting: It was a mix of introducing ourselves to them and introducing themselves to us, presenting and explaining our project from all it's sides, and orientation from them about what we will do during our sponsorship relation with them. They who suggested the idea of making a blog to our project in order to publish our work, if any one saw our blog and can help us in accomplishing our project, and to serve who will think about doing the same project in the future so as not to start from the beginning; but start from where we ended.

Mine Hunter team contacts

Team Members:

• Abd-Allah Hazaa Abd-Allah
Email: abdallh_gamal@yahoo.com

• Ahmad Galal Abd-El-Naiem
Email: medo6447@yahoo.com

• Ebtesam Fouad Sayed
Email: bosycat_love86@yahoo.com

• Mahmoud Hassan Abd-El-Aziz
Email: modi9hassan@yahoo.com

• Ramy Essam Hassan
Email: ramyessam19872005@yahoo.com
Cellular: +20-10-647-1695

• Samer Refaat Mahmoud
Email: samerrefaat@yahoo.com

Under the Supervision of:

• DR. Aliaa Youssif
Email: aliaay@helwan.edu.eg

Mine Hunter team contacts

Team Members:
• Abd-Allah Hazaa Abd-Allah
Email: abdallh_gamal@yahoo.com

• Ahmad Galal Abd-El-Naiem
Email: medo6447@yahoo.com

• Ebtesam Fouad Sayed
Email: bosycat_love86@yahoo.com

• Mahmoud Hassan Abd-El-Aziz
Email: modi9hassan@yahoo.com

• Ramy Essam Hassan
Email: ramyessam19872005@yahoo.com
Cellular: +20-10-647-1695

• Samer Refaat Mahmoud
Email: samerrefaat@yahoo.com

Under the Supervision of:
• DR. Aliaa Youssif
Email: aliaay@helwan.edu.eg

Sponsorship

- We made a great effort to get our currently sponsor, ITWorx, First trial was arranging a meeting with OMS, software development company, but we was at the beginning of the year, so we didn't have the ability to accomplish the whole research about our project idea, so that we couldn't make a powerful presentation and this situation had a big effect on the company's decision to sponsor us.

- Another trial to get a sponsor was with Link Dev, software development company also, we had a meeting with them but unfortunately they had a saturation with sponsoring teams, as they mentioned .

- After that we send E-Mails to a lot of companies; finally we got a response from ITWorx and for who doesn't know, it is a big reputable company in the field of software development, THANK GOD . OH I was about to forget, when we send our proposal to ITWorx it was ranked the fourth one among other eleven proposals.

Abstract

Mine clearing is a long and dangerous task. A lot of time is lost clearing areas free of mines that were selected just to be on the safe side. Deminers, such as Governments and Military organizations, are looking forward to methods that can reduce the suspected areas. There are two types of techniques to determine the location of land mines; Infrared Thermography(IR) and Ground Penetrating Radar (GPR). We are going to use GPR in the detection process as it can detect all types of mines, APM-ATM-UXO, and that it can discover mines at any depth

Motivations

- There are more than 100 million anti-personnel or anti-tank mines in more than 70 countries at this moment.
- It has been estimated that more than 26,000 people are killed or maimed by mines every year, which is one victim every 20 minutes. Some countries have severe mine casualties, and most of the victims are children.
- The cost to buy and lay a typical antipersonnel mine is between $3 and $30, while the cost to remove a mine is between $300 and $1,000. The European Commission and the United States have invested 138 million dollars for mine actions for the last two years, but this is just the tip of the iceberg considering the present clearance rate.
- As current mines do not have metallic material as much as older types, it is difficult to detect mines by the current employed technology, the metal detector (MD). Also, MD has been reported making too many false alarms in the former battlefield due to small fragments of munitions. Manual detection, called probing, works well for all kinds of mines, but the cost of labor and slow speed are encouraging the development of other techniques.
- Years of war have left millions of scattered and unrecorded landmines and unexploded ordnance (UXO) in scores of countries. The current nature of war and terrorism places the threat of landmines squarely on the doorstep of civilians. Civilian men, women and especially children, who often mistake mines and UXO for toys, make up the bulk of all mine accident victims in peacetime.
- Casualty Demographics As in previous years, in 2006 civilians accounted for three quarters of recorded casualties and children were 34 percent of civilian casualties, nearly all boys. In some severely affected countries/areas children were the majority of casualties (Afghanistan: 59 percent, Nepal: 53, Somaliland: 66) and boys between five and 14 years were a particularly high risk group. Males were 89 percent of all casualties where gender details were known; the gender and/or age of 1,454 people (25 percent of all casualties) were unknown. Some 24 percent of casualties were military; this increase from 2005 (19 percent) is due to one country, Colombia, which accounts for 57 percent of all military casualties. Excluding Colombia, 12 percent of casualties would be military.Other factors leading to recording of a higher military casualty rate are increased conflict (Pakistan) and extensive media reporting focused on foreign troops (Afghanistan and Iraq) at the expense of national civilian casualties.
• 1,205 casualties in 19 countries/areas in Sub-Saharan Africa, up from 1,122 casualties in 21 countries/areas in 2005.
• 2,510 casualties in 13 countries in the Asia-Pacific region, down from 3,031 in 16 countries/areas in 2005.
• 165 casualties in eight countries/areas in Europe, down from 335 in 10 countries/areas in 2005.
• 205 casualties in 11 countries/areas in the Commonwealth of Independent States, down from 228 in 11 countries/areas in 2005.
• 539 casualties in 13 countries/areas in the Middle East-North Africa, down from 990 in 12 countries/areas in 2005.
• 1,127 casualties in four countries in the Americas, down from 1,167 in eight countries in 2005.
• 14 countries/areas where casualties had occurred in 2005 had no casualties in 2006.
• Four countries with no casualties in 2005 had new casualties recorded in 2006: ???Republic of Congo (one), Hungary (one), Indonesia (five) and Tunisia (one)??
-Contributions by mine-affected countries/areas reported in this year’s Landmine Monitor country
reports include the following:
• Albania provided $233,000, in addition to funding of rehabilitation projects and unvalued in-kind contributions.
• Angola allocated $2.5 million for mine clearance in 2006, compared to $3 million in 2005.
• Azerbaijan provided $1.2 million in 2006, compared to roughly $750,000 in 2005 and $250,000 in 2004.
• Bosnia and Herzegovina contributed BAM20, 070,706 ($12.5 million) from central and local authorities, an increase from BAM17, 753,131 ($11.3 million) in 2005 (about 45 percent of the mine action budget in both years).
• Cambodia provided $1.2 million for mine action administration and programming.
• Chad contributed CFA165 million (some $300,000) to complement funding by UNDP.
• Chile provided $1.4 million, compared to approximately $1 million in government and armed forces contributions in 2005.
• Colombia provided COP2.562 billion ($1.1 million) for July 2006-June 2007, a large increase from $213,000 in July 2005-June 2006.
• Croatia provided HRK246, 757,250 ($42.3 million) or 82 percent of mine action funding from state budgets and state and local bodies, compared with HRK192, 769,625 ($32.4 million) or 57 percent in 2005.

Environment factors have a great effect on mines detection and extraction for that reason we will talk about those factors in detail, beside that we will see the different types of mines from the perspective of shapes, weight, and size.

There are some discrete variables that affect sensor performance. Here we will cover the common ones that affect the performance of both sensors (IR, and GPR):
1- Soil magnetic properties: The magnetic properties of the soil can strongly affect the received electromagnetic waves received in response from buried objects. This is a phenomenon verified in some type of soils, which are called magnetic soils. Variables taking this phenomenon into account could be magnetic susceptibility Ҳ of the soil, but interactions of magnetic properties with sensor measurements are still far to be reliably quantified by means of Ҳ. Therefore this variable is simply represented by a binary value (yes, no) standing respectively for highly magnetic soils and non-magnetic soils.
2- Soil moisture: The soil moisture influences the penetration depth of radar waves into the underground by attenuating them more quickly. Nevertheless, moderate level of soil humidity can improve images obtained from GPR sensors. The most suitable variable to take account of the soil moisture is the volumetric soil humidity percentage. Although it is a continuous value from 0% to 100%, its effect is evident for extremely dry or extremely saturated soils (more than 50 % humidity) by one side, and for moderately moist soils by the other.
3- Soil composition: This variable takes into account the fact that the particular materials (sand, clay, etc.) can enhance or attenuate thermal contrasts. The figure of merit of this variable will be a finite set of possible soil compositions, each one containing certain percentage of the main materials. Thus, a sandy soil will have 90% of sand and 10% of clay, for instance.
4- Soil uniformity: Soil uniformity takes into account the heterogeneity of underground materials composing the soil. Layers of different materials have different dielectric constants and this affects GPR measurements. Presence of soil non uniformities reflects in GPR images, which become hardly readable. Soil uniformity discrete variable has a binary value (yes, no), the former indicating presence of a multilayered heterogeneous soil, the latter a homogeneous soil.
5- Vegetation: Vegetation on soil is important for infrared images. In fact, it can cause non uniformity of solar heating, masking soil and mines signatures. It is therefore important to know if there is much vegetation over a soil zone.
6- Time of the day: The time of the day influences the contrast of thermal signatures. In fact, thermal anomalies are a diurnal cycle phenomenon. Following the time of the day it’s possible to have more or less evident signatures. Several trials have been made to establish a robust relation between hours of the day and maxima in temperature difference between bare soil temperature and above-mine soil temperature. Nevertheless, no precise and reliable results are available yet; therefore it is possible to give only wide range of hours representing levels of soil illumination, supposing operating hours for minefield clearance from 7 to 21.
7- Weather. Rain is an important factor for the thermal signatures. It reduces solar heating of the soil. A sunny weather favors solar heating and consequently image contrast. Prolonged time without solar heating will drive the necessity of active illumination. Weather condition is represented by a discrete variable assuming the values ‘clear’, ‘rain’ and ‘overcast’.

Classification of Mines

- The targets of this work, abandoned mines. We are looking for abandoned explosives, whose purpose has expired. According to their original purpose, they can be categorized into three types, antipersonnel mines (APM), antitank mines (ATM), and unexploded ordnance (UXO).
- Generally, UXO means misfired shells or bombs that should have been already exploded but still remain for some reason. UXO are usually found under the former battlefield. MD can easily detect them because they were not intentionally hidden. Many ATM include metallic material, and their size is bigger than the size of APM. Since they have been designed to destroy only vehicles, they are detonated by high pressure or the existence of a big metallic object. Thus, this demining work for ATM is relatively safer than the APM case. Also, detection is relatively easier because of their size and material.
- APM are most difficult to find and remove, and have damaged the most civilian victims. Most APM are made of nonmetallic material, and their size is much smaller than the size of ATM.
- The detonator is very sensitive to pressure. APM can be divided into three types; blast, bounding fragmentation, and directional fragmentation. The blast type mines are the most common targets for the humanitarian demining work. Their size is relatively smaller and lighter than other types. Sometimes they move by floating on rivers. They are usually buried underground, but some models can be installed by scattering from an airplane.
- Therefore, they can be found in any place, underground, on the surface, at a riverside. Since their mechanism is simple and their material is cheap, small military groups have been able to manufacture this type of mine. This fact has resulted in serious mine problems for some poor countries that cannot afford the investment for the demining work. The bounding fragmentation type mines are relatively bigger than the blast type, but they can destroy a larger area, while the blast type mines damage only the target within a limited distance.
- They are buried underground or fixed on the ground. Direct pressure or a trip wire activates their detonator. Once the trigger is activated, they bound up to some altitude and explode spreading out their fragments up to an area of 30m in radius, which is very lethal. Most directional fragmentation type mines are set up on the ground, and they spread their fragments in a specific direction. Some model’s lethal range is more than 200m. Since they are detonated by manual operation as well as a trip wire, sometimes this type of mine is considered an active weapon.