![]() ![]() ![]() However, real-time localization is an essential requiremen in implementing mobile robotics in agriculture. Moreover, the application of robotics in these tasks can have impact in the agricultural economic sector . The robot can be endowed with camera systems and artificial intelligence to learn what a trunk is. The vineyard context makes sense to provide the robot with the ability to recognize vine trunks as high-level features to use in the localization and mapping processes. However, the extraction of reliable and persistent features in an outdoor environment is a challenging task. Feature-based localization is one of the most common approaches to do so . For a robot to navigate safely in the vineyard, it needs to be able to localize itself. For mobile robots, the capability of autonomously navigating in steep slope vineyards has a mandatory requirement: real-time localization. Moreover, they can transform and have a significant impact on many agricultural economic sectors . These machines can be used to perform operations, such as planting, harvesting, monitoring, supply of water, and nutrients . The vast extension of the vineyards and their challenging conditions lead to an increasing need for human labor substitution by automatic and autonomous machines. In these cases, some theoretical assumptions were verified. Additional experiments were performed, where the impact of the amount of training data and the comparison between using Transfer Learning and training from scratch were evaluated. The experiments show that our trained models can detect trunks with an Average Precision up to 84.16% and our assisted annotation tool facilitates the annotation process, even in other areas of agriculture, such as orchards and forests. Finally, an assisted annotation tool was proposed to make the process of dataset building easier and improve models incrementally. Additionally, we deployed these models in an Edge-AI fashion and achieve high frame rate execution. VineSet was used to train state-of-the-art Single Shot Multibox Detector models. As significant contributions, we made available a novel vine trunk dataset, called VineSet, which was constituted by more than 9000 images and respective annotations for each trunk. Thus, Deep Learning models were trained and deployed to detect vine trunks. This work proposes the detection of semantic natural landmarks that are to be used in Simultaneous Localization and Mapping algorithms. All of these factors impose the implementation of precise and reliable navigation algorithms, so that robots can operate safely. An example of this is the mountain vineyards, built-in steep slope hills, which are characterized by satellite signal blockage, terrain irregularities, harsh ground inclinations, and others. Agricultural environments are particularly unstructured and, therefore, challenging to the implementation of robotics. The development of robotic solutions in unstructured environments brings several challenges, mainly in developing safe and reliable navigation solutions. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |