This project focuses its attention on the detection and quantification of Atrophic Rhinitis (A.R.), a disease prevalent in pigs. A.R. causes some of the bony structure in a pig's snout to atrophy, which exposes the animal to other diseases that may effect its growth. It is studied by examining a cross section of the pig's snout. By comparing areas and perimeters of the nasal cavity to bony structures, or turbinates in the snout, one can determine the severity of the atrophy. Currently, the veterinarians diagnosis is based on a subjective analysis of the snout. The problem with this method is that it varies among veterinarians. One would like to automate this procedure so that the system analyzes the same criteria (i.e. areas, perimeters) for all snouts. Semi-automated methods of manual planimetry are too slow and cumbersome to be of much use. The goal of this project is to develop a PC based system which could be installed in the slaughterhouse and could quickly determine the severity of Atrophic Rhinitis. This report describes the procedure and summarizes the results of software which has been developed to automate the task of determining the severity of A.R. Software has been developed to define the area of interest within the snout, create a histogram of the image, automatically threshold the gray level image, and compute areas and perimeters from the binary image.
Mussman, Timothy, "The Detection and Quantification of Atrophic Rhinitis Using Ditigal Image Processing Techniques" (1991). Opportunities for Undergraduate Research Experience Program (OURE). 144.