Skin Disease Diagnosis using Convolutional Neural Network -- 3
Budget: ₹1,500 – ₹12,500 INR
- Dataset used : contact me for the dataset link
- Input Image for diagnosis is taken from webcam
- Image preprocessing includes the following
> standardisation to a standard size (120*120)\
> Images are rotated by 90 degree in all directions and flipped to enhance dataset
> noise reduction (like hair and skin pigments)
> image enhancement(brings affected parts into focus by enhancing the area)
> segmentation into various segments to differentiate from normal skin
> feature extraction
- Softmax classifier is used for skin disease classification
- Convolutional Neural Network is used with initial batch size as 20 and epoch size as 25
- model is saved
- backend using jupyter lab and basic front end using .net framework to take webcam input and display prediction.
- primary focus on acne and Shingles while classifying others as well
- project is to be replicated as in reference paper attached
- Input Image for diagnosis is taken from webcam
- Image preprocessing includes the following
> standardisation to a standard size (120*120)\
> Images are rotated by 90 degree in all directions and flipped to enhance dataset
> noise reduction (like hair and skin pigments)
> image enhancement(brings affected parts into focus by enhancing the area)
> segmentation into various segments to differentiate from normal skin
> feature extraction
- Softmax classifier is used for skin disease classification
- Convolutional Neural Network is used with initial batch size as 20 and epoch size as 25
- model is saved
- backend using jupyter lab and basic front end using .net framework to take webcam input and display prediction.
- primary focus on acne and Shingles while classifying others as well
- project is to be replicated as in reference paper attached