AUTOMATED ELEPHANT TRACKER

Transcription

1 INTERNATIONAL JOURNAL OF ELECTRONICS AND COMMUNICATION ENGINEERING & TECHNOLOGY (IJECET) ISSN (Print) ISSN (Online) Volume 6, Issue 1, January (2015), pp IAEME: Journal Impact Factor (2014): (Calculated by GISI) IJECET I A E M E AUTOMATED ELEPHANT TRACKER Avirup Basu Electronics and Communication engineering, Siliguri Institute of Technology, Sukna, India, Aritro Mukherjee Electrical engineering, Siliguri Institute of Technology, Sukna, India, ABSTRACT In recent days the collisions between wildlife and railway have increased causing a very vital loss to the fauna. There have been huge numbers of deaths of elephants while crossing the railway tracks and the deaths have increased in the recent days and it has caused a serious problem. The main cause of death is there are no early warning systems which can give an early information regarding their location. Thus our main objective will be to provide an early warning system to the railway department so they can act accordingly. Our automated elephant tracker is a system developed with a motive of saving this elephants and to provide an early warning regarding their location. It consists of a manual bot capable of running on rough terrains. It tracks a fixed area but at the same time can also be manually controlled according to the circumstances. It has a wireless camera mounted on it which leads visualization of the elephants and provides early information to the base station regarding the locations and detections of this elephants. The main job of the base station is to control our elephant tracker and at the same time to process the signal to detect the location of elephant. Accordingly it sends the information to the railway authorities to take according steps. Which can minimize the collision and ensure the safety of the elephants and railway system. Keywords: Aforge, Arduino, C#, Euclidean, WPF. 1. INTRODUCTION Automated elephant tracker is a system developed with the motive to save elephants from deaths by speeding trains. The system is developed by incorporating 3 major systems. 1.1: Manual + Autonomous bot A manual bot named Black etrack v will compose the first major part of the system. The manual bot can move through rugged terrains near a railway track. It has two modes of 17

2 operation. Manual and automatic. The BET gets it power from a series of four 9V batteries. It attains mobility through 4 wide wheels. It receives signal from the base station and it transmits the signal to the base station by a wireless camera. The bot gets recharged from jumper stations (Charging points) at regular intervals. 1.2: Wireless A/V camera A wireless A/V camera is used to transfer the image from the field to the Base Station. The wireless camera captures the image and then it send it to the Base Station by the transmitter. The camera is relatively of 5x5 sq. cms. 1.3: Base Station A Base station is the key part to the system. The base station performs the set of tasks mentioned below in a hierarchal way. 1. Control the BET. 2. Receive the signal in video format from the camera 3. Process the video such that it can detect elephants using colour recognition and size recognition. 4. If detected send a signal to the train drivers to slow down. 2. FACTS AND FIGURES OF ELEPHANT DEATHS CAUSED BY SPEEDING TRAINS 2.1: Deadly Tracks The 21st century has seen a rapid growth of population like never before. India, the second largest populated country in the world, in particular has seen a rise in demand regarding economical and quick transportation systems. India's railway network being among the biggest in the world, is constantly spreading itself in order to increase the connectivity between remote locations thus, fulfilling the requirement of the teeming millions. But, while doing so, most of us fail to realize its impact on the biota. We fail to realize that we aren't alone on this planet & that there are millions of other species trying to make their living along with us. Instead of sharing our bread with them, we tend to be ignorant about them and are always on the run to prove ourselves as the most intelligent species on the planet. As a result of our in-human behaviour, these poor creatures (both big & small); hardly able to decipher the complexities of human nature, have to pay a heavy price, which may sometimes lead to their own death. With reference to the state's affidavit (state of West Bengal) there have been, 57 elephants have died in train hits in the past 19 years. In spite of repeated warnings and law enforcements, the railways have failed to prevent loss of lives in the elephant corridors, which is quite evident from the statistical data we see. 2.2: An excerpt from national daily, The Telegraph, 14 th November, 2013 A railway official today said that, the Jaipur-KamakhyaKabiguru Express, which mowed down five elephants in the Chapramari forest, was running at kmph, which is double or more the speed allowed when trains travel through forests. Since 2002, when the railway tracks between Siliguri Junction and Alipurduar were converted to broad gauge, 40 elephants have died in train accidents. A study by the WWF shows that most of the Human Elephant conflict occurs in the forest border areas. So our main area of concentration are the forest border areas. 3. HARDWARE MODULE The hardware of this system mainly consists of the bot. The bot is the main field player. It can be manually (remotely) controlled or automatically controlled. Based on it, the hardware inside the bot consists of the following parts. 18

3 Table 1: Parts List Si No Name Specifications Quantity 1 12V High Torque DC Geared Motor 300 RPM, High Torque 300RPM 4 2 Dual DC motor Driver Dual motor Driver 20A 2 4 Wireless, A/V camera 2.2Ghz 1 5 Arduino Leonardo Headers included 1 6 RF 2.4Ghz Serial link 2.4Ghz 1 8 RF 2.4Ghz USB link 2.4Ghz 1 9 Wheels 2cm width 10 cm diameter 4 The bot is controlled by RF network operating in 2.5 GHz frequency range. The robot is built on an open source multipurpose chassis. The chassis contains spots for 6 motors, 1 castor wheel, 1 line sensor array, and being an open source, it can be fitted with almost any type of components. The BET is driven by a 4 wheel drive and it follows a differential drive steering system. The differential drive steering system works on the principal of difference of motor speeds. 4. SOFTWARE MODULE The software is mainly divided into two units. The first unit comprises the controlling unit while the later comprises of the processing unit which deals with the image processing part. 4.1: Black e Track controller The Black e Track controller is the controlling unit of the bot. It sends control to the BET by serial port communication using a RF network of 2.4GHz. The signal on the other hand is interpreted by the receiver and the serial data is transferred to the Arduino Leonardo Board. The Arduino Leonardo Board sends the signal to the motor driver circuit to perform the desired operation. The Black e Track controller has 15 buttons. Si no Name of the control Corresponding control reflected in BET 1 Front_fast Makes the BET move forward at maximum speed 2 Back_fast Makes the BET move backward at maximum speed 3 Right_fast Makes the BET turn right at maximum speed 4 Left_fast Makes the BET turn left at maximum speed 5 Stop Makes the BET stop 6 Front_slow Makes the BET move forward at 60% of its original speed 7 Back_slow Makes the BET move backward at 60% of its original speed 8 Right_slow Makes the BET turn right at 65% of its original speed 9 Left_slow Makes the BET turn left 65% of its original speed 10 Connect Once you click the connect button, the controller will be connected to the wireless communicator s USB link 11 Disconnect Once you click the connect button, the controller will be disconnected to the wireless communicator s USB link deg right Makes the BET turn right continuously until and unless the stop button is pressed deg left Makes the BET turn left continuously until and unless the stop button is pressed 14 Press to activate keyboard controls This special control is used to shift the control of BET from a pointing input device to a Keyboard. 15 Activate This is a unique security button. The password that is entered just above the button is the activation code for the entire controller. Once entered, the user has to press the Activate button to activate the controls 19

4 The Black e Track Controller v is developed in Visual Studio 2013 using.net framework v in C# and xaml. It contains a status box where the current command is displayed. There are 3 progress bars. When the progress bar is in the intermediate stage, the corresponding set of commands is executed. In the controller, the Press to activate keyboard controls button activates the keyboard commands. The keyboard commands are mentioned in a tabular format below. Si No Keyboard keys Commands executed 1 W Fast forward 2 A Fast left turn 3 S Fast backward 4 D Fast right turn 5 Numpad 8 Slow forward 6 Numpad 2 Slow Backward 7 Numpad 4 Slow left turn 8 Numpad 6 Slow right turn Another button named Activate button is a special purpose security button. A field exist where you need to enter the password. By default the password is Unless you press the activate button after entering the password, all the controls go in the disabled stage. Only after entering the password and pressing the activate button, the controls can be enabled. Fig 1: Screenshot of Black e Track Controller V : Elephant detection system The Elephant detection system processor is the core component of the system. The EIDS receives signal from an on-board A/V wireless camera operating in 1-2 GHz frequency range. The signal is received by a receiver capable of interfacing various signals of varying signal range. The EIDS processor processes the image received to detect elephants. We have used a simple algorithm completely based on image color and image size to detect elephants. The algorithm utilizes an image processing [1] library called Aforge.NET [2]. The algorithm utilises two main classes. The first one is the Euclidean color extractor class and the second is the Blob class. 20

5 Euclidean color extractor filter: This filter is available in Aforge.NET framework. This filter utilises two major parameters to extract a certain color. The parameters includes the RGB values for the filters base color and the other parameter includes the size in pixels around the centre color. The filter filters pixels, which color is inside/outside of RGB sphere with specified centre and radius - it keeps pixels with colors inside/outside of the specified sphere and fills the rest with specified color. The filter accepts 24 and 32 bpp color images for processing. Fig 2: Image before applying the filter. Picture curtsey: On the above image as shown in Illustration 3, we apply the Euclidean Colouring filter with the centre color is of 215, 30, 30 values of Red, Green, Blue respectively and a colour sphere radius of 100 pixels. After applying the filter, we get the image as shown in Fig 3. Fig 3: Image after applying the filer. Note: only the areas with red color remains. Picture curtsey: Thus from the illustration 3, it is clear that the centre part of the rose image is filtered out. In the case of EIDS, the centre color is roughly the average color of an elephant. In the EIDS system, an option exist for the change of the centre color but it isn t there in EIDS v It will be available under v Please note that the Euclidean colouring filter is probably one of the most basic way to approach the detection of elephants. Blob class filter: The use of the blob class is simply to highlight the extracted region extracted by the Euclidean colouring filter. In EIDS v , there is an added facility where if a blob is displayed in the source image, then the word DETECTED is also displayed in red colour over the button right corner of the source image. 21

6 5: FUTURE SCOPE Fig 4: Screenshot of the EIDS processor v The Black e Track works well as a surveillance robot. The camera attached with the BET is digitally processed to detect elephants. Not only in elephant tracking, but also in other applications including military and other places where remotely operated vehicles are required. The BET can also be used in inspection of tunnels, pipelines and other potentially dangerous places for humans. Since BET uses an open source Arduino board, the BET can be re-programmed to perform other autonomous + manual jobs as well. Work is being carried out to convert the BET into a robot that can be wirelessly charged. To increase the flexibility of charging, the future version of BET will be incorporated with thin layer solar panel. The BET will have the capability to perform in both automated and manual modes of operation. Lots of security protocols are to be incorporated in the control unit of the Black e Track. We will also improve the EIDS processor to improve faster and accurate detection. REFERENCES Books 1. Digital Image Processing, Rafael C.Gonzalez, Richard E.Woods, and ISBN: Chandramouli.H, Dr. Somashekhar C Desai, K S Jagadeesh and Kashyap D Dhruve, Elephant Swarm Optimization for Wireless Sensor Networks A Cross Layer Mechanism International journal of Computer Engineering & Technology (IJCET), Volume 4, Issue 2, 2013, pp , ISSN Print: , ISSN Online: M.M.Kodabagi and Shridevi.B.Kembhavi, Recognition of Basic Kannada Characters In Scene Images Using Euclidean Distance Classifier International journal of Computer Engineering & Technology (IJCET), Volume 4, Issue 2, 2013, pp , ISSN Print: , ISSN Online: Manish P. Pujara, Lav Kumar and Ashish Mogra, Two Phase Flow Void Fraction Measurement Using Image Processing Technique International Journal of Mechanical Engineering & Technology (IJMET), Volume 4, Issue 3, 2013, pp , ISSN Print: , ISSN Online: Websites