Assessing optimal frequency for image acquisition in computer vision systems developed to monitor feeding behavior of group-housed Holstein heifers.

Computer vision systems have emerged as a potential tool to monitor the behavior of livestock animals. Such high-throughput systems can generate massive redundant data sets for training and inference, which can lead to higher computational and economic costs. The objectives of this study were (1) to develop a computer vision system to individually monitor detailed feeding behaviors of group-housed dairy heifers, and (2) to determine the optimal frequency of image acquisition to perform inference with minimal effect on feeding behavior prediction quality. Eight Holstein heifers (96 ± 6 d old) were housed in a group and a total of 25,214 images (1 image every second) were acquired using 1 RGB camera. A total of 2,209 images were selected and each animal in the image was labeled with its respective identification (1-8). The label was annotated only on animals that were at the feed bunk (head through the feed rail). From the labeled images, 1,392 were randomly selected to train a deep learning algorithm for object detection with YOLOv3 ("You Only Look Once" version 3) and 154 images were used for validation. An independent data set (testing set = 663 out of the 2,209 images) was used to test the algorithm. The average accuracy for identifying individual animals in the testing set was 96.0%, and for each individual heifer from 1 to 8 the accuracy was 99.2, 99.6, 99.2, 99.6, 99.6, 99.2, 99.4, and 99.6%, respectively. After identifying the animals at the feed bunk, we computed the following feeding behavior parameters: number of visits (NV), mean visit duration (MVD), mean interval between visits (MIBV), and feeding time (FT) for each heifer using a data set composed by 8,883 sequential images (1 image every second) from 4 time points. The coefficient of determination (R2) was 0.39, 0.78, 0.48, and 0.99, and the root mean square error (RMSE) were 12.3 (count), 0.78, 0.63, and 0.31 min for NV, MVD, MIBV, and FT, respectively, considering 1 image every second. When we moved from 1 image per second to 1 image every 5 (MIBV) or 10 (NV, MDV, and FT) s, the R2 observed were 0.55 (NV), 0.74 (MVD), 0.70 (MIBV), and 0.99 (FT); and the RMSE were 2.27 (NV, count), 0.38 min (MVD), 0.22 min (MIBV), and 0.44 min (FT). Our results indicate that computer vision systems can be used to individually identify group-housed Holstein heifers (overall accuracy = 99.4%). Based on individual identification, feeding behavior such as MVD, MIBV, and FT can be monitored with reasonable accuracy and precision. Regardless of the frequency for optimal image acquisition, our results suggested that longer time intervals of image acquisition would reduce data collecting and model inference while maintaining adequate predictive performance. However, we did not find an optimal time interval for all feeding behavior; instead, the optimal frequency of image acquisition is phenotype-specific. Overall, the best R2 and RMSE for NV, MDV, and FT were achieved using 1 image every 10 s, and for MIBV it was achieved using 1 image every 5 s, and in both cases model inference and data storage could be drastically reduced.

Características de fermentação da silagem de seis variedades de milho indicadas para a região semiárida brasileira / Fermentation parameters of silages of six maize varieties recommended for the Brazilian semi-arid region

Avaliaram-se as características fermentativas e a qualidade das silagens de seis variedades de milho, de ciclos precoce e superprecoce - BRS Caatingueiro, BRS Assum Preto, BR 5033 Asa Branca, BR 5028 São Francisco, Gurutuba e BRS 4103 - indicadas para a região semiárida brasileira. Foram utilizados silos experimentais, em delineamento inteiramente ao acaso, com seis tratamentos (variedades) e quatro repetições. Avaliaram-se: matéria seca (MS), matéria orgânica (MO), proteína bruta (PB), fibra em detergente neutro (FDN), fibra em detergente ácido (FDA), extrato etéreo (EE), carboidratos totais (CHO), carboidratos não fibrosos (CNF), pH, nitrogênio amoniacal como parte do nitrogênio total (N-NH3/NT), ácidos orgânicos e digestibilidade in vitro da matéria seca (DIVMS) das silagens. Os valores médios encontrados para a silagem foram: MS= 28,7 por cento; MO= 94,9 por cento; PB= 8,3 por cento; FDN= 49,9 por cento; FDA= 27,5 por cento; EE= 3,8 por cento; CHO= 82,7 por cento; CNF= 32,8 por cento; pH= 3,8; N-NH3/NT= 2,9 por cento/NT; ácido láctico = 7,6 por cento; ácido acético = 0,6 por cento; ácido butírico = 0,3 por cento e DIVMS= 57,9 por cento. As variedades BR 5028 - São Francisco e Gurutuba destacaram-se das demais em relação ao teor de matéria seca. A variedade BRS Caatingueiro apresentou maior teor de carboidratos não fibrosos em relação às demais. As silagens de todas as variedades foram classificadas como de excelente qualidade, por apresentarem potencial para ensilagem no semiárido brasileiro.

The fermentation characteristics and silage quality of six maize varieties of early and super early cycles were evaluated. They are recommended for the Brazilian semi-arid region (BRS Caatingueiro, BRS Assum Preto, BR 5033 - Asa Branca, BR 5028 - São Francisco, Gurutuba and BRS 4103). Experimental silos were used, in a completely randomized design, with six treatments (varieties) and four replicaties. The evaluated parameters were: dry matter (DM), organic matter (OM), crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), ether extract (EE), total carbohydrates (CHO), non-fibrous carbohydrates (NFC), pH, ammoniacal nitrogen as part of the total nitrogen (N-NH3/TN), organic acids, and in vitro dry matter digestibility (IVDMD) of the silages. The mean values found for silage were: DM= 29.6 percent; OM= 94.9 percent; CP= 8.2 percent; NDF= 49.9 percent; ADF= 27.5 percent; EE= 3.8 percent; CHO= 82.7 percent; NFC= 32.8 percent; pH= 3.8; N-NH3/TN= 2.9 percent/TN; lactic acid = 7.6 percent; acetic acid = 0.6 percent; butyric acid = 0.3 percent; and IVDMD = 57.9 percent. Varieties BR 5028 - São Francisco and Gurutuba stood out (P<0.05) from others in relation to dry matter. The BRS Caatingueiro showed higher (P<0.05) level of non-fiber carbohydrates in relation to the others. The silages from all the varieties were considered of excellent quality, with potential to be conserved as silage in the Brazilian semi-arid.