Automated classification of corn is important for corn sorting in intelligent agriculture. This paper presents a reliable corn classification method based on techniques of computer vision and machine learning. To discriminate different damaged types of corns, a line profile segmentation method is firstly used to segment and separate a group of touching corns. Then, twelve color features and five shape features are extracted for each individual corn object. Finally, a maximum likelihood estimator is trained to classify normal and damaged corns. To evaluate the performance of the proposed method, a private dataset consisting of images of normal corn and six kinds of damage corns, including heat-damaged, germ-damaged, cob-rot-damaged, blue eye mold-damaged, insect-damaged, and surface mold-damaged, were collected in this work. The proposed method achieved an accuracy of 96.67% for the classification between normal corns and the first four common damaged corns, and an accuracy of 74.76% was achieved for the classification between normal corns and six kinds of damaged corns. The experimental results demonstrated the effectiveness of the proposed corn classification system.

Weeding is very critical for agriculture due to its importance for reducing crop yield loss. Accurate recognition of weed species is one of the major challenges for achieving automatic and precise weeding. To improve the recognition performance of weeds and crops with similar visual characteristics, a fine-grained weed recognition method based on Swin Transformer and two-stage transfer learning is proposed in this study. First, the Swin Transformer network is introduced to learn the discriminative features that can distinguish subtle differences between visually similar weeds and crops. Second, a contrastive loss is applied to further enlarge the feature differences between different categories of weeds and crops. Finally, a two-stage transfer learning strategy is proposed to address the problem of insufficient training data and improve the accuracy of weed recognition. To evaluate the effectiveness of the proposed method, we constructed a private weed dataset (MWFI) with maize seedling and seven species of associated weeds that are collected in the farmland environment. The experimental results on this dataset show that the proposed method achieved the recognition accuracy, precision, recall, and F1 score of 99.18%, 99.33%, 99.11%, and 99.22%, respectively, which are superior to the performance of the state-of-the-art convolutional neural network (CNN)-based architectures including VGG-16, ResNet-50, DenseNet-121, SE-ResNet-50, and EfficientNetV2. Additionally, evaluation results on the public DeepWeeds dataset further demonstrate the effectiveness of the proposed method. This study can provide a reference for the design of automatic weed recognition systems.

This study used Silibinin as an additive to conduct fermentation experiments, wherein its effects on rumen gas production, fermentation, metabolites, and microbiome were analyzed in vitro. The silibinin inclusion level were 0 g/L (control group), 0.075 g/L, 0.15 g/L, 0.30 g/L, and 0.60 g/L (experimental group). Fermentation parameters, total gas production, carbon dioxide (CO₂), methane (CH₄), hydrogen (H₂), and their percentages were determined. Further analysis of the rumen microbiome’s relative abundance and α/β diversity was performed on the Illumina NovaSeq sequencing platform. Qualitative and quantitative metabolomics analyses were performed to analyze the differential metabolites and metabolic pathways based on non-targeted metabolomics. The result indicated that with an increasing dose of silibinin, there was a linear reduction in total gas production, CO₂, CH₄, H₂ and their respective percentages, and the acetic acid to propionic acid ratio. Concurrent with a linear increase in pH, when silibinin was added at 0.15 g/L and above, the total volatile fatty acid concentration decreased, the acetic acid molar ratio decreased, the propionic acid molar ratio increased, and dry matter digestibility decreased. At the same time, the relative abundance of Prevotella, Isotricha, Ophryoscolex, unclassified_Rotifera, Methanosphaera, Orpinomyces, and Neocallimastix in the rumen decreased after adding 0.60 g/L of silibinin. Simultaneously, the relative abundance of Succiniclasticum, NK4A214_group, Candidatus_Saccharimonas, and unclassified_Lachnospiraceae increased, altering the rumen species composition, community, and structure. Furthermore, it upregulated the ruminal metabolites, such as 2-Phenylacetamide, Phlorizin, Dalspinin, N6-(1,2-Dicarboxyethyl)-AMP, 5,6,7,8-Tetrahydromethanopterin, Flavin mononucleotide adenine dinucleotide reduced form (FMNH), Pyridoxine 5′-phosphate, Silibinin, and Beta-D-Fructose 6-phosphate, affecting phenylalanine metabolism, flavonoid biosynthesis, and folate biosynthesis pathways. In summary, adding silibinin can alter the rumen fermentation parameters and mitigate enteric methane production by regulating rumen microbiota and metabolites, which is important for developing novel rumen methane inhibitors.

Body condition score (BCS) of dairy cows is the direct reflection of their nutritional status. The timely estimation of BCS is beneficial to improving dairy cow health, milk production and reproduction. In this work, we propose an intelligent Edge-IoT platform with deep learning for estimating BCS of dairy cow, by integrating inference capability of deep learning and low latency of edge computing in IoT framework. Through capturing images of dairy cow’s back with the RGB-D camera, inference module deployed in the edge computing device firstly performs cow detection to localize the separate area of each dairy cow, and then performs individual identification and estimating BCS of dairy cows simultaneously. The existing systems are mainly commercial systems such as DeLaval and HerdVision, they use electronic ear tags with radio-frequency identification sensors for cow identification. Compared to existing systems, in the proposed platform, combined the finetuned YOLOv7 model and Avoid Repeated Inference (ARI) algorithm to detect dairy cow. An EfficientID model combined with metric learning is designed for cow identification, and an EfficientBCS model with Coordinate Attention (CA) is proposed for estimating BCS. The dairy cow’s identity (ID) and BCS are finally transmitted to the cloud analysis center. Experimental results show that the accuracy of estimating BCS reached 85% within 0.5 range error conducted on the test set collected in the dairy farm. The total inference time for one dairy cow is 3.138 seconds. Results show that the platform can be served as an excellent application of dairy cow body condition scoring.

Cow-YOLO: Automatic cow mounting detection based on non-local CSPDarknet53 and multiscale neck