Main Article Content

Abstract

Crops are vulnerable to wild animals. Therefore, it is very important to monitor the nearby presence of animals. Then the actuation of various devices should follow to repel the hazardous animals. Traditional methods have been widely applied depending on the kinds of produce and imperiling animals. In this paper, we propose a method to protect farms from wild animals via ubiquitous wired network devices, which is applied to farm along with traditional methods to improve the protection performance. Operational amplifier circuits are utilized mainly for the detection of animal intrusion from the outside of farms. The proposed monitoring scheme is to provide an early warning about possible intrusion and damage by wild animals. The performances of image recognition methods such as Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA) and Local Binary Patterns Histograms (LBPH) are tested and compared for the image recognition of the input animal images. The main idea of this paper is to present an independent, comparative study and some of the benefits and drawbacks of these most popular image recognition methods. Two sets of experiments are conducted for relative performance evaluations. In the first part of our experiments, the recognition accuracy of PCA, LDA, HoG and LBPH is demonstrated. Further, a method for combining the classifier into the Minimum Distance Classifier (i.e., Cascade classifier) is done which statistically guarantees the background extraction and pictures only the object of interest. The proposed system is tested with animal database and if the wild animals are detected, then the Alerted by Buzzer.

Article Details

How to Cite
Y.Keerthana, S. Purushothaman, S. Periyathambi, & P. Vanchi. (2018). A novel approach for animal recognition by using various recognition methods based on enhanced hybrid classifier technique. International Journal of Intellectual Advancements and Research in Engineering Computations, 6(2), 1057–1063. Retrieved from https://ijiarec.com/ijiarec/article/view/713