Quality Classification of Enoki Mushroom Caps Based on CNN

Agriculture is vital to human survival and remains one of the main driving forces of several economies in the world, and more so in developing economies. Agriculture is an important industry in China. With the increase of agricultural demand, it is urgent to reduce costs while maximizing agricultural production. As there are few traditional algorithms for enoki mushroom detection, this paper proposed a automatic enoki mushroom caps classification algorithm, and built a convolutional neural network model based on LeNet. The existing preprocessing approaches and network models based on convolutional neural network are improved and fine-tuned to realize the recognition of enoki mushroom caps. Experimental results demonstrate that CNN-driven classification application has higher recognition rate for enoki mushroom caps, which provides an important reference for the application of enoki mushrooms in agricultural automation production and helps to optimize yield and increase productivity.