Leaf nitrogen (N) directly correlates to chlorophyll production, affecting crop growth and yield. Farmers use soil plant analysis development (SPAD) devices to calculate the amount of chlorophyll present in plants. However, monitoring large-scale crops using SPAD is prohibitively time-consuming and demanding. This paper presents an unmanned aerial vehicle (UAV) solution for estimating leaf N content in rice crops, from multispectral imagery. Our contribution is twofold: (i) a novel trajectory control strategy to reduce the angular wind-induced perturbations that affect image sampling accuracy during UAV flight, and (ii) machine learning models to estimate the canopy N via vegetation indices (VIs) obtained from the aerial imagery. This approach integrates an image processing algorithm using the GrabCut segmentation method with a guided filtering refinement process, to calculate the VIs according to the plots of interest. Three machine learning methods based on multivariable linear regressions (MLR), support vector machines (SVM), and neural networks (NN), were applied and compared through the entire phonological cycle of the crop: vegetative (V), reproductive (R), and ripening (Ri). Correlations were obtained by comparing our methods against an assembled ground-truth of SPAD measurements. The higher N correlations were achieved with NN: $0.98$ (V), $0.94$ (R), and $0.89$ (Ri). We claim that the proposed UAV stabilization control algorithm significantly improves on the N-to-SPAD correlations by minimizing wind perturbations in real-time and reducing the need for offline image corrections.

Traditional methods to measure spatio-temporal variations in biomass rely on a labor-intensive destructive sampling of the crop. In this paper, we present a high-throughput phenotyping approach for the estimation of Above-Ground Biomass Dynamics (AGBD) using an unmanned aerial system. Multispectral imagery was acquired and processed by using the proposed segmentation method called GFKuts, that optimally labels the plot canopy based on a Gaussian mixture model, a Montecarlo based K-means, and a guided image filtering. Accurate plot segmentation results enabled the extraction of several canopy features associated with biomass yield. Machine learning algorithms were trained to estimate the AGBD according to the growth stages of the crop and the physiological response of two rice genotypes under lowland and upland production systems. Results report AGBD estimation correlations with an average of r = 0.95 and R² = 0.91 according to the experimental data. We compared our segmentation method against a traditional technique based on clustering. A comprehensive improvement of 13% in the biomass correlation was obtained thanks to the segmentation method proposed herein.

Traditional methods to measure spatio-temporal variations in above-ground biomass dynamics (AGBD) predominantly rely on the extraction of several vegetation-index features highly associated with AGBD variations through the phenological crop cycle. This work presents a comprehensive comparison between two different approaches for feature extraction for non-destructive biomass estimation using aerial multispectral imagery. The first method is called GFKuts, an approach that optimally labels the plot canopy based on a Gaussian mixture model, a Montecarlo-based K-means, and a guided image filtering for the extraction of canopy vegetation indices associated with biomass yield. The second method is based on a Graph-Based Data Fusion (GBF) approach that does not depend on calculating vegetation-index image reflectances. Both methods are experimentally tested and compared through rice growth stages: vegetative, reproductive, and ripening. Biomass estimation correlations are calculated and compared against an assembled ground-truth biomass measurements taken by destructive sampling. The proposed GBF-Sm-Bs approach outperformed competing methods by obtaining biomass estimation correlation of $0.995$ with $R^2=0.991$ and $RMSE=45.358$ g. This result increases the precision in the biomass estimation by around $62.43\%$ compared to previous works.

This paper addresses recent developments on Modular Snake Robots suited for applications on de-mining tasks. Inclusion of robotic platforms on the de-mining process will help tasks like the recognizance and perception of dangerous areas, minimizing human risk. As a mobile robot, the modular snake robot can move in a uneven terrain similar to a mined scenario. That is why, the main research efforts have been done on locomotion. Several contributions on this area are reported including gaits like: closed chain rolling, lateral rolling and a gait that use a helix shape both on the outside of cylinders. Another gait studied regards the swing movement of a modular robot chain hanging from one of its end modules. Contributions on robot architecture are reported here, including a distributed model predictive control and a multi-camera and sensor fusion. The perception problem of these robots also are researched by our lab and a pair of developments are shown here.

In this article we propose an algorithm for the classification, tracking and counting of vehicles and pedestrians in video sequences; The algorithm is divided into two parts, a classification algorithm, which is based on convolutional neural networks, implemented using the You Only Look Once (YOLO) method; and a proposed algorithm for tracking regions of interest based in a well defined taxonomy. For the first stage of classification, We train and evaluate the performance with a set of more than 50000 labels, which we make available for their use. The tracking algorithm is evaluated against manual counts in video sequences of different scenarios captured in the management center of the Secretaria distrital de Movilidad of Bogota.

Non-invasive crop phenotyping is essential for crop modeling, which relies on image processing techniques. This research presents a plant-scale vision system that can acquire multispectral plant data in agricultural fields. This paper proposes a sensory fusion method that uses three cameras, Two multispectral and a RGB depth camera. The sensory fusion method applies pattern recognition and statistical optimization to produce a single multispectral 3D image that combines thermal and near-infrared (NIR) images from crops. A multi-camera sensory fusion method incorporates five multispectral bands: three from the visible range and two from the non-visible range, namely NIR and mid-infrared. The object recognition method examines about 7000 features in each image and runs only once during calibration. The outcome of the sensory fusion process is a homographic transformation model that integrates multispectral and RGB data into a coherent 3D representation. This approach can handle occlusions, allowing an accurate extraction of crop features. The result is a 3D point cloud that contains thermal and NIR multispectral data that were initially obtained separately in 2D.