Performance Analysis of SKIN CANCER DETECTION
Implementation Plan:
———————————-
Step 1: Initially we load the input images from the HAM10000 dataset and ISIC 2020.
Step 2: Next we perform the Preprocessing process, In this process we perform normalization, resizing, and augmentation is applied to the input image.
Step 3: Next we perform the Segmentation process; In this Step we will implement the active contour segmentation approach for the segmentation process.
Step 4: Next we perform the Feature extraction step, In this step we extract the features such as convexity, circularity, irregularity index, textural patterns, color features, region of interest, etc… by using ResNet50 transfer learning technique.
Step 5: Next, we perform the Classification process, In this process we used the Capsule fusion and to optimize the model’s parameters we used Attention guide fusion.
Step 6: The proposed approach is validated using several performance metrics:
6.1: Accuracy
6.2: Sensitivity
6.3: Specificity
6.4: AUC-ROC
6.5: Recall
6.6: F1-score
============================================================================================================================================
Software Requirement:
———————————
1. Tool: Python-3.11.4
2. Operating System: Windows 10(64-bit)
—————————————————————————————————————————————————–
Note:-
———
We perform the EXISTING process based on the REFERENCE 1 Title: – Skin cancer detection from dermoscopic images using deep learning and fuzzy k-means clustering
