Wild Elephant Detection and Automated Alarming system

HackaDev “Impact Uva 2015” conducted by Dialog API idea mart with the collaboration of UNDP. It was a two days hackthon which is conducted to develop the innovative and creative ideas of the 15 teams. For that we had introduced group called “Tynet 5”. 

 

This project addresses the overall problem of wild life conflict with human settlements. Specifically, our project addresses common problems such as:

• Difficulty in communicating with wild life authority in time.
• Lack of effective methods of addressing tense conflict situations.
• Use of electric fence to prevent entering wild animals into settlement zones can be harmful for both sides.
• Lack of innovative and sophisticated tracking systems to inform and help communities predict potential animal attacks.

Elephants wander into settlements causing disturbances for inhabitations and Brutal and illegal use of tactics has fatal results on wild life and also for the eco-system.

To solve the all above mentioned problems we had developed the system call it as the Bee-Phant.

 

Project objectives include: 

1). Develop computer based program to recognize wild animals (elephants) that have been a threatened for human settlements for implementing this objective here system uses camera to capture videos of wild animals near to the human settlements and by analyzing those videos recognize wild animals like elephants using digital image processing. And also track the human settlements that are threatening at that time.

2). Informing related authorities and ordinary people (villagers). After recognizing wild animals predict the attack and the system activate the alarming systems in threatened villages while authorize people will be notified the predicted attack. At the same time system will send notification for people through communication network (message system)through these methods most of the threatened people can react positively to the situation in advance

3). Automate the system to prevent wildlife from entering settlement zones .Recent researches show that elephants scared on buzz sound of bees. As animal react to the situation in tsunami they feel infrared frequency which cannot capture by human. So we take it as opportunity to develop effective tactics rather than electrical fence or Brutal to keep away wild animals that have been a threatened for human settlements and plantation.

Initially the video is captured and recognizing wild animals is also included in that method. Then the possible attack is predicted according to the identified animal and tracking the area that will be threatened is done. Then notification are sent to the wild life authority and to the inhabitants and at the same time the alarming system also activated. Simultaneously the automated system that prevents the entrance of wild animals to settlement zones is activated and keeping it active until the threat is away.

Using above use case diagram can easily understand the system. To develop this system we use .Net framework. To train the wild animals (Elephants) use Haar cascading technology in Image processing. To send the messages to the authorized bodies we use Ozeki SMS gateway.

 

Our system is really accurate and efficient. It only detects wild elephants and there is no chance to make alarm for other animals. Below there are several images of our project.
 

The green colored box tells that the elephants were detected.

 

 
  
when there are no any wild elephants system does not detect it as a threat.

 

Udara Rathnayaka
Roshan Gunarathna
Dilan Maduranga
Dulanjalee Lakmali
Neelamani hewawasam are the group members of Tynet 5 team