Classifying Animal Species from Footprint Images using Deep Learning Algorithm

  • Unique Paper ID: 166495
  • Volume: 11
  • Issue: 2
  • PageNo: 1158-1165
  • Abstract:
  • Footprint-based animal species classification is vital for wildlife monitoring and conservation. This study proposes an efficient method utilizing Probabilistic Neural Networks (PNNs) for classifying animal species from footprint images. The approach involves preprocessing the images to normalize and reduce noise, followed by extracting key features such as shape, texture, and Histogram of Oriented Gradients (HOG). The PNN, known for its pattern recognition prowess, processes these features to classify the footprints. The network’s structure, consisting of input, pattern, summation, and output layers, enables it to estimate the probability density functions of the input data effectively. Training the PNN with a labeled dataset of diverse footprint images showed high classification accuracy, outperforming traditional methods. This automated, non-invasive technique offers a scalable solution for accurate species identification, enhancing wildlife management efforts. Future work aims to broaden the dataset, incorporate hybrid models, and implement real-time applications for field use. This approach promises significant improvements in the efficiency and reliability of animal species classification based on footprints. The issues involved in developing an effective and successful system for tracking, recognizing, and classifying animals are the focus of this research project. Building algorithmic models for animal tracking, segmentation, detection, and classification has been attempted and accomplished with success. We propose two methods to effectively isolate an animal from its environment to aid in the effective taxonomic classification of species. The recommended animal segmentation method is assessed using performance indicators based on areas. We also present a classification model that utilises many features and classifiers. Among the different elements that are extracted from the segmented animal photographs are colour, gabor, and LBP.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 2
  • PageNo: 1158-1165

Classifying Animal Species from Footprint Images using Deep Learning Algorithm

Related Articles