A Novel Approach For Automatic Identification Of Psoriasis Affected Skin Area.

Abstract

The paper presents the detection of psoriasis affected skin. Thecolor feature of the skin from RGB model, where the skin color is defined as redness, greenness, and blueness are analyzed. An algorithm is developed to differentiate affected area from non affected area. The color histogram analysis is carried out for more than fifty samples to analyse erythema in order to differentiate affected skin from non affected skin. Today dermatologists visually analyse the patients for the line of treatment which is biased by various external factors. The proposed model for diagnosis is not subjective where the decisions are based on various external factors such as emotions, part of the day and vary from dermatologist to dermatologist which may have a great impact on the treatment of the disorder. The algorithm is objective and minimizes the deviation in the line of treatment as it is not affected by intra and inter diagnosis by dermatologists. The RGB histogram is analyzed and a model is built based on mean and standard deviation to differentiate healthy skin and psoriasis disorder affected skin.