首页|Institute for Systems and Computer Engineering Reports Findings in Robotics (Dee p learning based approach for actinidia flower detection and gender assessment)
Institute for Systems and Computer Engineering Reports Findings in Robotics (Dee p learning based approach for actinidia flower detection and gender assessment)
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-New research on Robotics is the subjec t of a report. According to news originating from Porto, Portugal, by NewsRx cor respondents, research stated, "Pollination is critical for crop development, esp ecially those essential for subsistence. This study addresses the pollination ch allenges faced by Actinidia, a dioecious plant characterized by female and male flowers on separate plants." Our news journalists obtained a quote from the research from Institute for Syste ms and Computer Engineering, "Despite the high protein content of pollen, the ab sence of nectar in kiwifruit flowers poses difficulties in attracting pollinator s. Consequently, there is a growing interest in using artificial intelligence an d robotic solutions to enable pollination even in unfavourable conditions. These robotic solutions must be able to accurately detect flowers and discern their g enders for precise pollination operations. Specifically, upon identifying female Actinidia flowers, the robotic system should approach the stigma to release pol len, while male Actinidia flowers should target the anthers to collect pollen. W e identified two primary research gaps: (1) the lack of gender-based flower dete ction methods and (2) the underutilisation of contemporary deep learning models in this domain. To address these gaps, we evaluated the performance of four pret rained models (YOLOv8, YOLOv5, RT-DETR and DETR) in detecting and determining th e gender of Actinidia flowers. We outlined a comprehensive methodology and devel oped a dataset of manually annotated flowers categorized into two classes based on gender. Our evaluation utilised k-fold cross-validation to rigorously test mo del performance across diverse subsets of the dataset, addressing the limitation s of conventional data splitting methods. DETR provided the most balanced overal l performance, achieving precision, recall, F1 score and mAP of 89% , 97%, 93% and 94%, respectively, highli ghting its robustness in managing complex detection tasks under varying conditio ns."
NETL
NSTL
