Recognition of Feline Epidermal Disease using Raspberry-Pi based Gray Level Co-occurrence Matrix and Support Vector Machine
Bryan James Andujar, Nica Jo Ferranco, Jocelyn F. Villaverde · 2021 IEEE 13th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management (HNICEM) · 2021
As of 2018, with over 373 million population of cats worldwide, cases of feline skin diseases have drastically increased as well, with 6-15% of feline patients who have experienced at least one form of dermatopathy in their lifetimes. Research on detecting feline skin diseases has been focused on using different diagnostic methods in the past. Examples of these methods include the Fur Pluck Method, which uses microscopic hair evaluation to detect parasites. Another study uses Wood’s Lamp method that uses a UV light source to detect dermatophytosis. In this study, the image processing technique in diagnosing two types of skin diseases, Dermatophytosis and Ectoparasitic Skin Disease, was identified using GLCM and SVM to extract the features and classify the images, respectively. The system trained 270 images with 90 photos each on Dermatophytosis, Ectoparasitic, and unknown skin diseases and tested 45 skin disease images. These feline skin disease images used are primarily from Philippine short-haired cats or Puspins. These feline skin disease images used are primarily from domestic short-haired cats. Confusion Matrix was used in determining the accuracy of the system. The accuracy of the system reached 86.776% with 80% on Dermatophytosis Skin Diseases, 93.33% for Ectoparasitic Skin Diseases, and 87% for the Unknown Parameter with 13.224% Error of Commission.