Improvement of mushroom image classification algorithm based on lightweight MobileViT network
Mushrooms are a group of fungi with wide distribution and diverse forms,including both edible and poisonous va-rieties.Accurate classification and identification of mushrooms are of great significance for ecological conservation.In recent years,deep learning has made remarkable achievements in the field of image classification.However,due to the small size and diversity of mushroom image datasets,traditional deep neural networks may face challenges such as high model complexity and computa-tional cost in mushroom image classification tasks.To address these issues,a lightweight network based on improved MobileViT is proposed.The network structure incorporates channel attention mechanism and fuses local and global features.Additionally,the 3×3 convolutional kernel in the Fusion module is replaced with a 1×1 convolutional kernel.Experimental results demonstrate that the improved model achieves better mushroom classification performance,with an average recognition accuracy of 87.5%.Com-pared to the original model and other neural network models,the accuracy is significantly improved.
万方数据
