A Frame Work for Detection of the Degree of Skin Disease Using Different Soft Computing Model
| Autor: | Manisha Barman, S. Biswas, J. Paul Choudhury |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Soft computing
Artificial neural network Computer science business.industry k-means clustering Image processing Pattern recognition 01 natural sciences Fuzzy logic 030218 nuclear medicine & medical imaging Image (mathematics) 010309 optics Euclidean distance 03 medical and health sciences Digital image 0302 clinical medicine 0103 physical sciences Genetic algorithm Artificial intelligence business |
| Zdroj: | Communications in Computer and Information Science ISBN: 9789811385773 CICBA (1) |
| DOI: | 10.1007/978-981-13-8578-0_5 |
| Popis: | Health Informatics plays a great role in the recovery of disease. In health informatics system, the information regarding the conditions of various diseases and subsequent remedy regarding application of medicines have been stored in Computer memory. Based on the conditions of sickness of a particular organ of the patient, the status of the disease of that person can be ascertained using the stored information in the computer and accordingly certain measures are advised by the doctor to the patients with the help of a computer. In this paper skin disease detection technique is proposed using image processing techniques and soft computing models. The soft computing models include Fuzzy logic, Artificial Neural Network and Genetic algorithm. Initially image characteristics have been collected from digital images using image processing tools. Soft computing models viz using Fuzzy logic, Artificial Neural Network and Genetic algorithm have been applied on the image characteristics. Based on the minimum value of average error, the particular soft computing model has been selected to produce the estimated value of image characteristics. K means clustering algorithm has been applied to the estimated value of image characteristics to produce optimal number of clusters. Each cluster represents the degree of the disease. Then image processing tool has been applied to the unknown image of a patient to get image characteristics of new image. Now the selected soft computing tool has been applied on the image characteristics of new image to produce the estimated image characteristics. The estimated image characteristics has been compared with the cluster centre value of the clusters as formed based on Euclidian distance function. The cluster centre with minimum distance function value indicates the degree of disease of a patient. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|