Nakhasys: An ML-Based Disease Diagnosing Application

Literatrure Review And Existing System

This section is used to present the survey or background methods of systems which are based on nail image analysis for different disease detection are given below.

• •

  Hardik and Shah (2012) proposed a method for extracting a section of a known image using color processing and palm image were used for analysis and experimental work. Noriaki et. al. in 2013 used fingernail detection using hand images including pal and distribution density, color continuity has been used to improve accuracy by Mente and Marulkar (2017).

• •

  Sharma and Shrivastava (2015) worked on nail color and texture and segmented images on the basis of texture and further used for analysis on the basis of color and texture.

• •

  In 2016, Gandhat et al. proposed the method for analysis of nail images using Haar Transforms Matrix to generate features. Row mean method has been used to reduce feature vector and then compare a query feature vector stored in template dataset using similarity measures like MSE, absolute difference. And then they find out matching and calculate GAR (Genuine acceptance ratio) and produced result.

• •

  The early stage disease detection system which process nail image has been proposed by Indi and Gunge in 2016 to extract the feature from sample data. The Weka tool has been used for training dataset using patients nail images. Using color detection algorithm and decision tree by C4.5 used for experimental work and achieved 65% results which correctly matched with training data set by Indi and Gunge (2016) and Singal and Arora (2015).

• •

  Saranya and Ranichitra (2017) used the image segmentation technique to detect nail abnormality. Median and average filters were used for preprocessing and converted into gray scale to improve the efficiency. Watershade, Thresholding and k-means segmentation techniques have been used to identify the nail shape and will be used for diagnosis purposed.

• •

  Thakare, Meshram, and Baradkar (2017) presented the medical support system for detecting human health conditions using Decision tree, Neural network and support vector machine. Based on nail color and texture analysis Bicolor algorithm and GLCM algorithm was also used.