首页|Researchers at University of Picardie Jules Verne Have Reported New Data on Robo tics and Automation (Focusing On Object Extremities for Tree Instance Segmentati on In Forest Environments)
Researchers at University of Picardie Jules Verne Have Reported New Data on Robo tics and Automation (Focusing On Object Extremities for Tree Instance Segmentati on In Forest Environments)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News Daily News – Data detailed on Robotics - Robotics a nd Automation have been presented. According to news reporting originating from Amiens, France, by NewsRx correspondents, research stated, “As part of the devel opment of many robotic systems for the forestry sector, forest scene understandi ng requires the use of computer vision algorithms. However, this dense and unstr uctured environment is complex and puts conventional detection approaches to the test.” Financial support for this research came from Agence Nationale de la Recherche ( ANR). Our news editors obtained a quote from the research from the University of Picar die Jules Verne, “In the case of tree instance segmentation, the presence of clo sely spaced or even intertwined trees, their highly variable shapes, and complex masks due to their branches and leaves are just some of the challenges to be ov ercome. For this, specific learning of tree boundaries is required to better dis tinguish one from another. In this letter, we propose ConvexMask, a convolutiona l neural network for real-time instance segmentation. ConvexMask opts for a labe l representation approach with a convex exterior polygon, defined by tree extrem ities, and a binary mask to handle the detail and occlusions that the label may contain.”
AmiensFranceEuropeRobotics and Aut omationRoboticsUniversity of Picardie Jules Verne
NETL
NSTL
