Real Time Animal Repellent System using Image Processing

In forest zone and agricultural field human animal conflict is a major problem where enormous amount of resources is lost and human life is in danger. In recent times the numbers of these kinds of conflicts are increasing. So this zone is to be monitored continuously to prevent entry of this kind of animals or any other unwanted intrusion. In our proposed project we have designed a system which detects motion using PIR sensors. After detection of motion the camera in that area is triggered to take an image of the intrusion. Then the image is sent to the image processing processor and algorithm to classify the kind of animal or human is applied. The algorithm used here is Content Based Image Classification which gives the result as to which animal has intruded and in some cases if it be a false alarm no alert is given and all further processes are not done. If detected as a threat the location and type of danger intruded is sent as an SMS to the field owner and also to the forest officials as a alert message using GSM technology. Also animal repellent techniques such as irritating noise in the range of ultrasonic frequencies and Bright ,light which repels most of the animals are used .This is a way of making the animal uncomfortable so that it retreats back from the agricultural field and also it would be easy for the humans to scare them away.