No abstract

The use of Unmanned Aerial Vehicle (UAV) can contribute to find solutions and add value to several agricultural problems, favoring thus productivity, better quality control processes and flexible farm management. In addition, the strategies that allow the acquisition and analysis of data from agricultural environments can help optimize current practices such as crop counting. The present research proposes a methodology based on the exploitation of deep learning approach, especially Convolutional Neural Networks (CNN) on UAV data for fruit tree detection and counting. We build models for the automatic extraction of fruit trees. This approach is divided into main phases: dataset pre-treatment, implementing a fruit trees detection model by exploiting several CNN architectures, validating and comparing the performances of different models. The exploitation of RGB UAV images as input information will allow the learning models to find a statistical structure, which will result in rules capable of automating the detection task. They can be applied to new images for automatically identify and count fruit trees. The application of the methodology on collected data has made it possible to reach estimates of detection and counting until 96 %.

Before the COVID-19 crisis, the Southern African Developing Countries (SADC) had a varied energy mix including renewable energy, fossil fuels, and military energy production. The use of fossil fuels in the energy mix is known to be the source of the growing levels of greenhouse gases in the atmosphere. However, there was a reduction in GHG emissions following the pandemic, which reduced travel and trade, and worldwide disruption in economic activities. The priority of priority B in the 2015-2020 Regional Indicative Strategic Development Plan, which is Energy, continues. As a result, the availability of affordable and renewable energy is still a priority for south of the equator countries and their growth agenda. This paper is aimed at exploring the sustainability of SADC countries' electricity sectors by using three sustainability pillars: Social, Environmental and Economic (SEE). SEE offers the main concepts of renewable energy, in a way that is socially, environmentally appropriate and economically viable. Study shows a gap in access rate in SADC countries with only Mauritius and Seychelles reaching 100% access to modern energy services (electricity) for both rural and urban areas. Currently all the member countries have set their RE goals for the year 2030. However, the subsidies by SADC member countries indicate that they are practiced as a way to make electricity affordable, and also to make electricity available to lower income households. In the period 2014-2017, big national budget deficits happened in various Southern African countries because of subsidies. Thus, this paper is of crucial importance to the foundational advancement of sustainable electricity sector growth in the country. The findings of this paper play a crucial role in helping and guiding politicians to better understand the existing and challenges future in the energy market and alternatives to address these problems. Additional research is given on how to arrive at sustainable decisions for the electricity sector in the region.