Authors :
Samiksha Dhote; Prajakta Dhumal; Prajwal Gaidhani; Indrajeet Ghadge; .S.R.Nalamwar
Volume/Issue :
Volume 7 - 2022, Issue 11 - November
Google Scholar :
https://bit.ly/3IIfn9N
DOI :
https://doi.org/10.5281/zenodo.7439981
Abstract :
Deep learning and image processing techniques
for skin disease identification are part of the suggested
solution. Skin conditions are brought on by a variety of
reasons, including DNA sequencing, radiation, and
mutations, which result in skin defects. If skin conditions
are not treated in a timely manner, they often spread to
other parts of the body. In order to be treated, skin
disorders must therefore be found in their early stages.
These symptoms need to be identified early because skin
disorders have been linked to mortality problems, lengthy,
expensive therapies, and numerous characteristics.
We are creating web application for early stage skin
disease detection. In the suggested solution, skin illnesses
will be identified from the provided image collection using
image processing techniques which uses a color image's
inputs. Depending on the training dataset, we have used the
Convolutional Neural Network which has an excellent
visual representation power for the recognition or detection
task. Therefore in order to diagnose skin diseases early with
more accuracy and efficiency transfer learning will be used
. It will provide the best training to the model and give a
high accuracy, precision, recall, specificity for the correct
anticipation of skin diseases among all picked which help
doctor for early detection and prevent chronic disorder as
well as economic and mental loss
Keywords :
Data Preprocessing, Outlier Detection, Image Classification, Deep Learning, Convolution Neural Network, Transfer Learning.
Deep learning and image processing techniques
for skin disease identification are part of the suggested
solution. Skin conditions are brought on by a variety of
reasons, including DNA sequencing, radiation, and
mutations, which result in skin defects. If skin conditions
are not treated in a timely manner, they often spread to
other parts of the body. In order to be treated, skin
disorders must therefore be found in their early stages.
These symptoms need to be identified early because skin
disorders have been linked to mortality problems, lengthy,
expensive therapies, and numerous characteristics.
We are creating web application for early stage skin
disease detection. In the suggested solution, skin illnesses
will be identified from the provided image collection using
image processing techniques which uses a color image's
inputs. Depending on the training dataset, we have used the
Convolutional Neural Network which has an excellent
visual representation power for the recognition or detection
task. Therefore in order to diagnose skin diseases early with
more accuracy and efficiency transfer learning will be used
. It will provide the best training to the model and give a
high accuracy, precision, recall, specificity for the correct
anticipation of skin diseases among all picked which help
doctor for early detection and prevent chronic disorder as
well as economic and mental loss
Keywords :
Data Preprocessing, Outlier Detection, Image Classification, Deep Learning, Convolution Neural Network, Transfer Learning.
