Research on Automatic and Efficient Digital Image Recognition Technology for Citrus Pest Based on Neural Network

A convolutional neural network (CNN) expert system has been developed to address the significant challenges posed by pests and diseases on "Citrus King" navel oranges, which are highly demanded in both domestic and international markets. Traditional detection methods for these afflictions are slow, subjective, and costly, necessitating professional expertise. Utilizing artificial intelligence and digital image technology, our CNN is capable of intelligently analyzing real-time images of navel oranges. It can automatically and swiftly identify various types of pests and diseases, such as Huanglongbing, scab, anthracnose, and fruit cracking disease. This facilitates early prevention and reduces the incidence of pests and diseases in navel orange cultivation. Moreover, this method potentially eliminates or minimizes pesticide use, thereby lowering planting costs. Experimental results have shown that the CNN’s accuracy in identifying navel orange pests and diseases exceeds 90%, which is a significant improvement over traditional methods, and it also substantially reduces detection costs.