One of the most dangerous forms of cancer, skin cancer has been on the rise over the past ten years. Nonetheless, melanoma detection is a method that uses deep learning algorithms to analyze images and accurately diag...One of the most dangerous forms of cancer, skin cancer has been on the rise over the past ten years. Nonetheless, melanoma detection is a method that uses deep learning algorithms to analyze images and accurately diagnose melanoma. An improved result for cancer treatment may result from early diagnosis. Then, in a matter of seconds, it will be simple to identify skin cancer using deep learning. In this research, a deep learning-based automatic skin cancer detection method is proposed. Data was considered from the ISIC database dataset which has 2357 images. To obtain average color information and normalize all color channel information, we used a few preprocessing approaches. Next, data was collected for categorization and reshaping of the images. To avoid overfitting, we additionally employed data augmentation. In the end, the Convolutional Neural Network was used to achieve our goal, which improved the accuracy of prediction. Using the Resnet50 algorithm, the accuracy rate rose to 98%, which will be helpful to get a good outcome with better accuracy.展开更多
文摘One of the most dangerous forms of cancer, skin cancer has been on the rise over the past ten years. Nonetheless, melanoma detection is a method that uses deep learning algorithms to analyze images and accurately diagnose melanoma. An improved result for cancer treatment may result from early diagnosis. Then, in a matter of seconds, it will be simple to identify skin cancer using deep learning. In this research, a deep learning-based automatic skin cancer detection method is proposed. Data was considered from the ISIC database dataset which has 2357 images. To obtain average color information and normalize all color channel information, we used a few preprocessing approaches. Next, data was collected for categorization and reshaping of the images. To avoid overfitting, we additionally employed data augmentation. In the end, the Convolutional Neural Network was used to achieve our goal, which improved the accuracy of prediction. Using the Resnet50 algorithm, the accuracy rate rose to 98%, which will be helpful to get a good outcome with better accuracy.
