Author(s): Lanka Dhana Lakshmi, Neelapala Siva Rama Krishna Babu, Mullapudi Sameera, Rangisetty Narendra Sai and Pasumarthi Sri Ram Pavan

Abstract: Because of their great number of species, visual similarities, and environmental consequences on picture pleasant, mushroom categorisation affords a primary problem in machine learning. using a set of pre-educated deep learning models-VGG16, MobileNet, MobileNetV2, ResNet50, InceptionV3, Xception, DenseNet169, ResNet101, LeNet, EfficientNetV2S, CNN-this paintings indicates a complete method the usage of transfer learning. The project intends to create a robust system capable to differentiate edible from inedible mushrooms depending on consumer-enter photos. Model performance is evaluated using accuracy, precision, recall, F1-rating amongst different criteria. Testing with real-world mushroom pix the use of a user-pleasant UI allows for immediate feedback on the categorisation findings. The work expects to attain excellent classification accuracy and dependability by leveraging transfer learning and a broad ensemble of neural network designs. This study facilitates to improve automated mushroom popularity systems, thereby perhaps improving safety and pride for mushroom aficioners all around.

DOI: 10.33545/27076571.2025.v6.i1a.134

Pages: 61-65 | Views: 1055 | Downloads: 511

Download Full Article: Click Here