Cloud/edge computing for compliance in the Brazilian livestock supply chain.

Brazil is an important player in the global agribusiness markets, in which grain and beef make up the majority of exports. Barriers to access more valuable sustainable markets emerge from the lack of adequate compliance in supply chains. Here is depicted a mobile application based on cloud/edge computing for the livestock supply chain to circumvent that limitation. The application, called BovChain, is a peer-to-peer (P2P) network connecting landowners and slaughterhouses. The objective of the application is twofold. Firstly, it maximizes sustainable business by reducing transaction costs and by strengthening ties between state-authorized stakeholders. Secondly, it creates metadata useful for digital certification by exploiting CMOS and GPS sensor technologies embedded in low-cost smartphones. Successful declarative transactions in the digital space are recorded as metadata, and the corresponding big data might be valuable for the certification of livestock origin and traceability for sustainability compliance in 'glocal' beef markets.

Counting Cattle in UAV Images-Dealing with Clustered Animals and Animal/Background Contrast Changes.

The management of livestock in extensive production systems may be challenging, especially in large areas. Using Unmanned Aerial Vehicles (UAVs) to collect images from the area of interest is quickly becoming a viable alternative, but suitable algorithms for extraction of relevant information from the images are still rare. This article proposes a method for counting cattle which combines a deep learning model for rough animal location, color space manipulation to increase contrast between animals and background, mathematical morphology to isolate the animals and infer the number of individuals in clustered groups, and image matching to take into account image overlap. Using Nelore and Canchim breeds as a case study, the proposed approach yields accuracies over 90% under a wide variety of conditions and backgrounds.

A Study on the Detection of Cattle in UAV Images Using Deep Learning.

Unmanned aerial vehicles (UAVs) are being increasingly viewed as valuable tools to aid the management of farms. This kind of technology can be particularly useful in the context of extensive cattle farming, as production areas tend to be expansive and animals tend to be more loosely monitored. With the advent of deep learning, and convolutional neural networks (CNNs) in particular, extracting relevant information from aerial images has become more effective. Despite the technological advancements in drone, imaging and machine learning technologies, the application of UAVs for cattle monitoring is far from being thoroughly studied, with many research gaps still remaining. In this context, the objectives of this study were threefold: (1) to determine the highest possible accuracy that could be achieved in the detection of animals of the Canchim breed, which is visually similar to the Nelore breed (Bos taurus indicus); (2) to determine the ideal ground sample distance (GSD) for animal detection; (3) to determine the most accurate CNN architecture for this specific problem. The experiments involved 1853 images containing 8629 samples of animals, and 15 different CNN architectures were tested. A total of 900 models were trained (15 CNN architectures × 3 spacial resolutions × 2 datasets × 10-fold cross validation), allowing for a deep analysis of the several aspects that impact the detection of cattle using aerial images captured using UAVs. Results revealed that many CNN architectures are robust enough to reliably detect animals in aerial images even under far from ideal conditions, indicating the viability of using UAVs for cattle monitoring.

Estresse por deficiência hídrica em genótipos de Brachiaria brizantha / Hydric deficiency in genotypes of Brachiaria brizantha

O experimento foi desenvolvido em casa de vegetação na Embrapa Pecuária Sudeste, utilizando acessos de Brachiaria brizantha. O delineamento experimental foi o de blocos ao acaso, com arranjo fatorial 2 X 4, sendo dois tratamentos (com e sem déficit hídrico) e quatro genótipos. O grupo de genótipos avaliado foi formado pelas cultivares 'BRS Piatã', 'Marandú', 'Xaraés' e 'BRS Paiaguás', sendo o experimento realizado durante o mês de julho de 2008. As variáveis analisadas foram massa seca total, massa seca de parte aérea, massa seca de raiz, massa seca de colmo, massa seca de folha, taxa de alongamento foliar, área foliar, área foliar específica. O estresse por deficiência hídrica exerceu efeito negativo em todas as características estudadas e em todos os acessos avaliados. A cultivar 'BRS Piatã' foi o genótipo que menos apresentou alteração entre os tratamentos com e sem déficit hídrico, indicando portanto, tolerância dessa cultivar em relação ao estresse por falta de água no solo nessas condições experimentais.

The experiment was conducted in a greenhouse at Embrapa Southeast Livestock, using Brachiaria brizantha accessions. The experimental design was fully randomized blocks with a 2 x 4 factorial arrangement, with two treatments (with and without water deficit) and four genotypes. The group evaluated was formed by genotypes 'BRS Piatã', 'Marandu', 'Xaraés' and 'BRS Paiaguás'. The experiment was conducted during the month of July 2008. The variables analyzed were dry mass of total plant, shoot mass, leaves, stems and roots, leaf elongation rate, leaf area and specific leaf area. The water deficit stress exerted negative effect on all characteristics studied and in all accessions. 'BRS Piatã' was the genotype that showed less change between treatments with and without water deficit, thus indicating tolerance of this cultivar in relation to water stress in the soil under these experimental conditions.

Simple agrometeorological models for estimating Guineagrass yield in Southeast Brazil.

The objective of this work was to develop and evaluate agrometeorological models to simulate the production of Guineagrass. For this purpose, we used forage yield from 54 growing periods between December 2004-January 2007 and April 2010-March 2012 in irrigated and non-irrigated pastures in São Carlos, São Paulo state, Brazil (latitude 21°57'42″ S, longitude 47°50'28″ W and altitude 860 m). Initially we performed linear regressions between the agrometeorological variables and the average dry matter accumulation rate for irrigated conditions. Then we determined the effect of soil water availability on the relative forage yield considering irrigated and non-irrigated pastures, by means of segmented linear regression among water balance and relative production variables (dry matter accumulation rates with and without irrigation). The models generated were evaluated with independent data related to 21 growing periods without irrigation in the same location, from eight growing periods in 2000 and 13 growing periods between December 2004-January 2007 and April 2010-March 2012. The results obtained show the satisfactory predictive capacity of the agrometeorological models under irrigated conditions based on univariate regression (mean temperature, minimum temperature and potential evapotranspiration or degreedays) or multivariate regression. The response of irrigation on production was well correlated with the climatological water balance variables (ratio between actual and potential evapotranspiration or between actual and maximum soil water storage). The models that performed best for estimating Guineagrass yield without irrigation were based on minimum temperature corrected by relative soil water storage, determined by the ratio between the actual soil water storage and the soil water holding capacity.irrigation in the same location, in 2000, 2010 and 2011. The results obtained show the satisfactory predictive capacity of the agrometeorological models under irrigated conditions based on univariate regression (mean temperature, potential evapotranspiration or degree-days) or multivariate regression. The response of irrigation on production was well correlated with the climatological water balance variables (ratio between actual and potential evapotranspiration or between actual and maximum soil water storage). The models that performed best for estimating Guineagrass yield without irrigation were based on degree-days corrected by the water deficit factor.