Corn is one of the global staple grains and the largest grain crop in China. During harvesting, grain separation, and corn production, corn is susceptible to mechanical damage including surface cracks, internal cracks, and breakage. However, the internal cracks are difficult to observe. In this study, hyperspectral imaging was used to detect mechanical damage in corn seeds. The corn seeds were divided into four categories that included intact, broken, internally cracked, and surface-crackedtv. This study compared three feature extraction methods, including principal component analysis (PCA), kernel PCA (KPCA), and factor analysis (FA), as well as a joint feature extraction method consisting of a combination of these methods. The dimensionality reduction results of the three methods (FA + KPCA, KPCA + FA, and PCA + FA) were combined to form a new combined dataset and improve the classification. We then compared the effects of six classification models (ResNet, ShuffleNet-V2, MobileNet-V3, ResNeSt, EfficientNet-V2, and MobileNet-V4) and proposed a ResNeSt_E network based on the ResNeSt and efficient multi-scale attention modules. The accuracy of ResNeSt_E reached 99.0%, and this was 0.4% higher than that of EfficientNet-V2 and 0.7% higher than that of ResNeSt. Additionally, the number of parameters and memory requirements were reduced and the frames per second were improved. We compared two dimensionality reduction methods: KPCA + FA and PCA + FA. The classification accuracies of the two methods were the same; however, PCA + FA was much more efficient than KPCA + FA and was more suitable for practical detection. The ResNeSt_E network could detect both internal and surface cracks in corn seeds, making it suitable for mobile terminal applications. The results demonstrated that detecting mechanical damage in corn seeds using hyperspectral images was possible. This study provides a reference for mechanical damage detection methods for corn.

Ammonia (NH₃) is a common agricultural gas, and its accurate detection is critical to agricultural production. In this study, nano-CuO/CeO₂ composites were prepared to achieve a wide range of ammonia detection at room temperature. Characterization data verified the composite heterojunction structure of CuO/CeO₂, which demonstrates an outstanding large specific surface area for ammonia detection. It provides more active sites for NH₃ molecules, which brings a very high response to ammonia (70.3% @100 ppm NH₃), a large detection range (0.5–200 ppm NH₃), and a fast response/recovery time (13 s/17 s @20 ppm NH₃). Systematic testing showed that the nano-CuO/CeO₂ composites also exhibit excellent extended-term stability and selectivity. Further studies showed that the p-n heterojunction structure of CuO/CeO₂ allowed the composite to retain its gas-sensitive properties to ammonia, in addition to the improved ammonia-detection range of the composite based on the synergistic effect of these two materials. The mechanism of CuO/CeO₂ heterojunction nanocomposites towards ammonia detection was also elucidated from a microscopic perspective at the molecular level. Finally, a triboelectric nanogenerator (TENG) that can be driven by wind power has been prepared, upon which the feasibility of the combination of the TENG and the ammonia sensor to realize environmental monitoring was investigated.

Rapid and non-destructive prediction of crude protein content in fresh mulberry leaves is essential for precision sericulture and feed quality management. In this study, a portable near-infrared spectrometer was used to collect spectra (900–1700 nm) from 80 mulberry leaf samples, and the predictive performance of partial least squares (PLS) and several deep learning models was systematically evaluated. The results demonstrate that deep learning models significantly outperform traditional PLS approaches. Notably, the proposed Chunk-wise Fully Connected Network (CFCN) achieves the best predictive performance (R²ₚ = 0.9869, RMSEP = 0.3044%, RPD = 8.75) with the fewest parameters among all compared models. Furthermore, the CFCN substantially reduces computational cost while maintaining superior accuracy over more complex hybrid architectures. These findings confirm that portable near-infrared spectroscopy combined with the CFCN enables high-precision, real-time crude protein prediction in mulberry leaves, offering a practical and cost-effective solution for on-site quality assessment and smart nutritional management in sericulture.