<p><strong>Abstract.</strong> Recently, recurrent neural networks have been proposed for crop mapping from multitemporal remote sensing data. Most of these proposals have been designed and tested in temperate regions, where a single harvest per season is the rule. In tropical regions, the favorable climate and local agricultural practices, such as crop rotation, result in more complex spatio-temporal dynamics, where the single harvest per season assumption does not hold. In this context, a demand arises for methods capable of recognizing agricultural crops at multiple dates along the multitemporal sequence. In the present work, we propose to adapt two recurrent neural networks, originally conceived for single harvest per season, for multidate crop recognition. In addition, we propose a novel multidate approach based on bidirectional fully convolutional recurrent neural networks. These three architectures were evaluated on public Sentinel-1 data sets from two tropical regions in Brazil. In our experiments, all methods achieved state-of-the-art accuracies with a clear superiority of the proposed architecture. It outperformed its counterparts in up to 3.8% and 7.4%, in terms of per-month overall accuracy, and it was the best performing method in terms of F1-score for most crops and dates on both regions.</p>

Abstract-Small animal PETICT devices provide anatomical and molecular imaging at the same time, enabling the joint visualization and analysis of both modalities. An accurate PET/CT alignment is required to correctly interpret these studies. A proper calibration procedure is essential for small animal imaging, since resolution is much higher than in human devices. This work presents an alignment phantom and a method that enable a reliable and replicable measurement of the geometrical relationship between PET and CT modules. The phantom can be built with laboratory materials, and is used to estimate the rigid spatial transformation that aligns both modalities. It consists of three glass capillaries located in noncoplanar triangular geometry and filled with FDG, so they are easily identified in both modalities. The method is based on automatic line detection and localization of the corresponding points between the lines on both modalities, which allows calculating the rigid alignment parameters. Different geometric configurations of the phantom (i.e. different angles and distances between capillaries) were tested to assess the repeatability of the calculations. To measure the alignment precision achieved, we attached two additional sodium point sources to the phantom, which were neglected in the registration process. Our results show that the accuracy of the alignment estimation, measured as average misalignment of the Na sources, is below half the PET resolution. The alternative settings for the phantom layout did not affect this result, indicating the low dependency of the alignment calculated with the actual phantom layout. Our approach allows measuring the PETICT transformation parameters using an in-house built phantom and with low computational effort and high accuracy, demonstrating that the proposed phantom is suitable for alignment calibration of dual modality systems on a real environment.

Abstract-IBASPM software is an atlas-based method for automatic segmentation of brain structures, available as a freeware toolbox for the SPM package. To test the influence of the atlas when segmenting normal and pathologic brains, manual segmentation of the caudate nucleus head was compared to automatic segmentations using four different atlases: the default MNI AAL atlas; a customized atlas created from a combined sample of patients (n=20) and controls (n=18); and a customized atlas obtained separately for each group. Maximum average ratio of overlapping voxels (dice overlap) between manual and automatic segmentation was 71 o~for controls and 52% for patients. In both groups, overlap ratios were better when using the customized atlases, instead of the standard MNI AAL atlas. Accuracy of the method was biased between left and right hemispheres, and also between groups, individual variability being higher in patients than in controls. Volumetric measurements using the customized atlases were also more accurate than using the MNI AAL atlas. Volume data were closer to manual segmentation values than dice overlap ratio (average differences ranging from 22.7°~for MNI AAL atlas to 10.1 % for customized atlas of patients and controls combined). Results suggests a low overaU performance of IBASPM as an automatic segmentation method for the head of the caudate nucleus. Because of the biases observed, the use of this method for analyzing caudate nucleus in patients presenting anatomical abnormalities should be cautiously carried out.

Abstract. Applying remote sensing technology to map and monitor agriculture and its impacts can greatly contribute for the proper development of this activity, promoting efficient food, fiber and energy production. For that, not only remote sensing images are needed, but also ground truth information, which is a key factor for the development and improvement of methodologies using remote sensing data. While a variety of images are current available, inclusive cost-free images, field reference data is scarcer. For agricultural applications, especially in tropical regions such as Brazil, where the agriculture is very dynamic and diverse (recent agricultural frontiers, crop rotations, multiple cropping systems, several management practices, etc.), and cultivated over a vast territory, this task is not trivial. One way of boosting the researches in agricultural remote sensing is to stimulate people to share their data, and to foster different groups to use the same dataset, so distinct methods can be properly compared. In this context, our group created the LEM Benchmark Database (a project funded by the ISPRS Scientific Initiative project - 2017) from the Luiz Eduardo Magalhães (LEM) municipality, Bahia State, Brazil. The database contains a set of pre-processed multitemporal satellite images (Landsat-8/OLI, Sentinel-2/MSI and SAR band-C Sentinel-1) and shapefiles of agricultural fields with their correspondent monthly land use classes, covering the period of one Brazilian crop year (2017–2018). In this paper we present the first results obtained with this database.

Abstract. Recent works have studied crop recognition in regions with highly complex spatio-temporal dynamics typical of a tropical climate. However, most proposals have only been evaluated in a single agricultural year, and their capabilities to generalize to dates outside the temporal sequence have not been properly addressed thus far. This work assesses the generalization capabilities of a recent convolutional recurrent architecture, testing it in a temporal sequence two years ahead of the sequence with which it was trained. Furthermore, a N-to-1 variant of such network is proposed, which is able to produce classification outcomes for every month in the agricultural year, and it is compared with two baselines designed in a more traditional approach, in which a separate specific network is trained for each month of the year. The approaches are evaluated on two public datasets from a tropical region. The first dataset comprehends the period from June 2017 to May 2018, while the second goes from October 2019 to September 2020. Results show a decrease of up to 24.6% in per-date average F1 score when training the network with data of an agricultural year different from the one it is tested on, which indicates a domain shift that demands further research. Additionally, the proposed approach presented only a slight decrease in performance compared to its baseline when trained on the same dataset, with a 2.7% drop in average F1 score. This performance drop is a small cost in exchange for its operational advantages, such as reduced training time and a more straightforward pipeline.