Multispectral images allow the identification of crop conditions that can be used for proper decision making in agriculture activities. A multispectral imaging system that includes cameras mounted on an Unmanned Aerial Vehicle (UAV) with other electronic modules such as GPS and autopilot system has been developed for covering great extensions of land and the acquisition of geolocated images. However, image resolution limits the altitude at which the UAV can fly and the image acquisition rate. Therefore, an integral design among the embedded system, digital cameras and trigger interface must be developed in order to guarantee an acceptable performance. This paper provides a design proposition for a multispectral system for interfacing with an onboard flight controller to synchronize data acquisition and obtain aerial images.
Climate change is affecting the agricultural production in Ancash - Peru and corn is one of the most important crops of the region. It is essential to constantly monitor grain yields and generate statistic models in order to evaluate how climate change will affect food security. The present study proposes as a proof of concept to use Deep learning techniques for the classification of near infrared images, acquired by an Unmanned Aerial Vehicle (UAV), in order to estimate areas of corn, for food security purpose. The results show that using a well balanced (altitudes, seasons, regions) database during the acquisition process improves the performance of a trained system, therefore facing crop classification from a variable and difficult-to-access geography.
The design, construction, and field testing of an unmanned aircraft system to monitor high altitude volcanoes is described in this paper. The system must be able to fly above the crater of the volcano and acquire important information such as images, video, and concentration of gases in order to assess potential risks to nearby population. The system must be able to fly above 6000 meters above sea level autonomously with telemetry data transmitted up to 7 km away. All aspects of the system were developed considering the extreme conditions at which the aircraft was subjected. Field experiments were accomplished on two volcanoes in Peru, being one of them currently under eruption.
Crop mapping is a vital tool for agricultural management and food security that can benefit from remote sensing data. The purpose of this research is to use machine learning (ML) techniques to classify quinoa crops from multispectral images. The spectral reflectance of five optical bands is used to develop classification models that are tested for diverse quinoa phenological phases. Deep learning methods Segnet and Unet were investigated, as well as Decision Trees, Discriminant Analysis, K nearest Neighbor, Support Vector Machines, Adaboost and Random Forest. Data was collected from quinoa crop fields in Cabana, Puno region in Peru. The multispectral images were captured using an Unmanned Aircraft System (UAS) from a height of 50 meters. Deep learning methods leave behind other approaches in the classification job, according to the results.
A remote system for the teaching and learning of digital electronics design is proposed. It is part of a formal engineering design course of digital circuits fundamentals designed to support student training during the pandemic due to closure of university campuses. Students can have access to lab equipment through the internet. The system uses a VPN connection to access remote computers in laboratories with FPGA DE10-Lite boards. Each board is connected to an Arduino Due board to gain accessibility to the I/O signals that are necessary for implementing digital circuits. With the use of digital design software Quartus Lite students develop designs on the FPGA Board. A Matlab application sends the control signals to the Arduino Due board to show the operation of the FPGA DE10-Lite by means of a webcam.
An automatic monitoring system for the detection of larvae of Peruvian Scallop in microscopic images is presented. Image processing algorithms and a statistic procedure have been applied in order to detect and count larvae contained in a microscopic image as well as to measure individual larva area.