Publikasjoner
NIBIOs ansatte publiserer flere hundre vitenskapelige artikler og forskningsrapporter hvert år. Her finner du referanser og lenker til publikasjoner og andre forsknings- og formidlingsaktiviteter. Samlingen oppdateres løpende med både nytt og historisk materiale. For mer informasjon om NIBIOs publikasjoner, besøk NIBIOs bibliotek.
2025
Forfattere
Krzysztof KusnierekSammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Sammendrag
Single-class object detection, which focuses on identifying, counting, and tracking a specific animal species, plays a vital role in optimizing farm operations. However, dense occlusion among individuals in group activity scenarios remains a major challenge. To address this, we propose YOLO-SDD, a dense detection network designed for single-class densely populated scenarios. First, we introduce a Wavelet-Enhanced Convolution (WEConv) to improve feature extraction under dense occlusion. Following this, we propose an occlusion perception attention mechanism (OPAM), which further enhances the model’s ability to recognize occluded targets by simultaneously leveraging low-level detailed features and high-level semantic features, helping the model better handle occlusion scenarios. Lastly, a Lightweight Shared Head (LS Head) is incorporated and specifically optimized for single-class dense detection tasks, enhancing efficiency while maintaining high detection accuracy. Experimental results on the ChickenFlow dataset, which we developed specifically for broiler detection, show that the n, s, and m variants of YOLO-SDD achieve AP50:95 improvements of 2.18%, 2.13%, and 1.62% over YOLOv8n, YOLOv8s, and YOLOv8m, respectively. In addition, our model surpasses the detection performance of the latest real-time detector, YOLOv11. YOLO-SDD also achieves state-of-the-art performance on the publicly available GooseDetect and SheepCounter datasets, confirming its superior detection capability in crowded livestock settings. YOLO-SDD’s high efficiency enables automated livestock tracking and counting in dense conditions, providing a robust solution for precision livestock farming.
Sammendrag
In broiler breeding, precise counting is crucial for improving production efficiency and ensuring animal welfare. Nevertheless, counting chickens precisely is a challenging task especially when young chicks always huddle for warmth. Although deep learning has been widely taken in different counting related tasks, more accurate localization and counting of chickens in high stocking density scenes still has not been well investigated. We propose a point supervised dense chickens flock counting network (PCCNet), which directly utilizes points as learning targets. The network adopts information feature fusion to assist the identification of broilers high stocking density scenes. In addition, considering the distance of neighboring points as matching cost in point matching algorithms is advantageous for generating more reasonable matching results, facilitating model convergence. To validate the effectiveness of the proposed network, a Chicken Counting Dataset (CCD) is built, consisting of two subsets separated by different ages: CCD_A and CCD_B. The accuracies of PCCNet on the two subsets of CCD are 97.85% and 97.06%, with corresponding Mean Absolute Errors (MAE) of 1.966 and 5.173, and Root Mean Square Errors (RMSE) values of 3.474 and 7.034, respectively. Our model achieves better broiler counting performance than other state-of-the-art (SOTA) methods.
2024
Forfattere
Helene Leirset TysdalSammendrag
In a world with war, hunger and increasing expenses, food production is an important aspect. Every country has a need and a responcibility to be able to produce enough food for its own people. Wheat is an important crop that can contribute to increased self sufficiency in Norway. The Norwegian milling industry largely controls the proportion of domestic wheat to be used for milling. Because of that they have imported a lot of the wheat needed to produce flour in the past. In the cases where Norwegian wheat have been used, it has mostly been spring wheat. Winter wheat has a high yield potential, and can ve important to utilize to achive goals for increased self-sufficienct fot the Norwegian agriculture, but more research is required to meet the milling industries requirements. Winter wheat production may be an important step in this process. This master thesis aimes to reachers how split nitrogen fertilication affects yield components and protein in the grain. As well as how N-sensor measurments at heading and flowering can be an effective tool to guide the farmers. Five fertilization strategies, four locations and tre variaties was used to reachers the effect of split nitrogen fertlization. Each location had three repetitions and every repition included every possible combination of fertilization strategy and variaty. Plant nitrogen level was monitored during the growing season to predict the protein content at harvest. The results showed that for protein percentage, the fertilization regiment is the most important aspect, hereunder total nitrogen content and the distributio between the different applications. Giving 8 kg of nitrogen at the first application and 6 kg of nitrogen at the second application gave a significantly higher protein content compered to giving 14 kg of nitrogen at the first application and nothing at the second. To get a higher yield the most important aspect is the winter wheat variety and Kuban gave a significantly higher yield than Bernstein and Ellvis. N-sensor measurments and measurments of the real nitrogen uptake was done at heading and flowering. The N-sensor measurmetns underestimated the nitrogenuptake especially in field with drought. Regression analysis found that 2/3 of the observed variation could be explained by the model at z49 and z65. More research and a bigger dataset is thereby needed to get a better model.
Sammendrag
Raman spectroscopy is a powerful and non-invasive analytical method for determining the chemical composition and molecular structure of a wide range of materials, including complex biological tissues. However, the captured signals typically suffer from interferences manifested as noise and baseline, which need to be removed for successful data analysis. Effective baseline correction is critical in quantitative analysis, as it may impact peak signature derivation. Current baseline correction methods can be labor-intensive and may require extensive parameter adjustment depending on the input spectrum characteristics. In contrast, deep learning-based baseline correction models trained across various materials, offer a promising and more versatile alternative. This study reports an approach to manually identify the ground-truth baselines for eight different biological materials through extensively tuning the parameters of three classical baseline correction methods, Modified Multi- Polynomial Fit (Modpoly), Improved Modified Multi-Polynomial Fitting (IModpoly), and Adaptive Iteratively Reweighted Penalized Least Squares (airPLS), and combining the outputs to best fit the training data. We designed a one-dimensional Transformer (1dTrans) tailored to fit Raman spectral data for estimating their baselines, and evaluated its performance against convolutional neural network (CNN), ResUNet, and three aforementioned parametric methods. The 1dTrans model achieved lower mean absolute error (MAE) and spectral angle mapper (SAM) scores when compared to the other methods in both development and evaluation of the manually labeled original raw Raman spectra, highlighting the effectiveness of the method in Raman spectra pre-processing.
Sammendrag
Det er ikke registrert sammendrag
Forfattere
Therese With Berge Jiangsan Zhao Jakob Geipel El Houssein Chouaib Harik Malin Larsen Græsdahl Marit HelgheimSammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag
Sammendrag
Det er ikke registrert sammendrag