Improved CNN Method for Crop Pest Identification Based on Transfer Learning.

Timely treatment and elimination of diseases and pests can effectively improve the yield and quality of crops, but the current identification methods are difficult to achieve efficient and accurate research and analysis of diseases and pests. To solve this problem, this study proposes a crop pest identification method based on a multilayer network model. First, the method provides a reliable sample dataset for the recognition model through image data enhancement and other operations; the corresponding pest image recognition and analysis model is constructed based on VGG16 and Inception-ResNet-v2 transfer learning network to ensure the completeness of the recognition and analysis model; then, using the idea of an integrated algorithm, the two improved CNN series pest image recognition and analysis models are effectively fused to improve the accuracy of the model for crop pest recognition and classification.

A danger-theory-based immune network optimization algorithm.

Existing artificial immune optimization algorithms reflect a number of shortcomings, such as premature convergence and poor local search ability. This paper proposes a danger-theory-based immune network optimization algorithm, named dt-aiNet. The danger theory emphasizes that danger signals generated from changes of environments will guide different levels of immune responses, and the areas around danger signals are called danger zones.