Research

11 Feb, 2019 |Facebook |Twitter |Email |


Integral Professors develops a system towards Wildlife safety.

Human-wildlife Conflict is one of the major threats to Indian wildlife, human activities such as deforestation, Habitat loss, Lack of prey and illegal roads cut through the forest have threatened the safety and survival of wildlife in India. Speeding vehicles on various roads are killing many animals annually, mostly Chital Deer, mouse deer, Fox, birds, snakes and nocturnal animals such as Indian civets, hare and mouse deer and some time the big cats including tiger and leopard and Elephants.
Two of Integral University Professors have come together with their student, to find a way out for the safety of these animals while they cross the roads and communicate with the drivers, to alert them of any wild sightings.

“Elephant and tiger both are in the extinct stage. The protection of these animals is the priority of forest officials. The government has planned various ways to protect both animals. But still, animals are killed in various ways. There are many highways and railway tracks going through the forest which are curved and thus prevent the drivers from seeing the animal on the roads or tracks thus drivers are unable to stop their vehicles and animals get killed” shares Dr. Preetam Suman, Assistant Professor , Computer Science & Engineering.
Dr. Suman is an expert in Image Processing and Machine Learning and has developed devices for the security of forest and wildlife. He has worked in Panna Tiger Reserve for protection of Tigers from poachers. He has developed an invisible sensor using optical fiber cable and deployed in the forest to monitor the illegal movement.

In the continuation of project “Protection of forest and wildlife”, Dr. Preetam and Ms. Roshan Jahan along with Til Theunissen, from Aachen University, Germany, under DAAD RISE worldwide research internship program, have developed an algorithm with higher accuracy for the identification of tiger/elephant and a mechanism to communicate with drivers.

The focus of the project is to implement a system, which should be able to recognize the presence of elephant/tiger near the railway tracks and roads and inform vehicle/train driver to slow down or stop the vehicle.

The Research team further explains that animals, especially Elephants come near tracks and get killed by trains and vehicles because of their speed. The algorithm is combining image processing operations like background subtraction, template matching and color analysis involving many steps.

  • The first step is to enhance the image and make the scene stable under different (weather) conditions. For example, rain appears as a tiny change in pixel data and will be erased by smoothing the image.
  • The second step is to subtract the background to get only moving objects. Small objects like leaves were removed by applying morphological operations.
  • The third step is template matching which found essential parts for identification like legs, size of the envelope, and the aspect ratio of the bounding box.
  • The fourth step is to find the mean color of the object is an additional characteristic for the identification of elephants/tigers.

A database of images is used for testing and gives results of about 80 percent accuracy, which makes the algorithm interesting for further development and improvement.


Integral Times is a publication of Integral University Media & Communications.

<