Severity level prediction of acne using hybrid MOA-FCM segmentation algorithm with ANN classifier

A person needs self-confidence in order to act normal as well as desired manner. Physical state of an individual, which sorts feelings of inadequacy and melancholy, which is one of the internal elements contributing to lack of confidence in adolescents. Adolescents who suffer from acne frequently experience a persistent skin condition that causes inflammation and/or obstruction of their hair follicles. Dermatologists manually count a number of pimples to determine their severity when treating acne. Doctor must exert a great deal of effort as well as this procedure has a significant degree of unreliability with inaccuracy. To avoid manual prediction of skin disease, a computer-assisted image processing method to acne identification as well as classification is proposed. This models initial step is gathering data, which consists of facial images with acne. The second process consists of image resizing, Lucy–Richardson deconvolution and contrast stretching for pre-processing the image and the third process is segmentation of the pre-processed image using the hybrid MOA-FCM algorithm. The features of the segmented images are extracted by using the grey-level co-occurrence matrix and these features are trained and tested by using the ANN. The proposed model's performance measures are contrasted with those of existing methods. The evaluated values of accuracy, precision, specificity and error for the proposed model are 0.97, 0.96, 0.98 and 0.027. Thus, the proposed severity level prediction of acne using a hybrid MOA-FCM segmentation algorithm with ANN classifier model performs better compared to the existing techniques.