Scholarlink

<p>Accurate localization is a fundamental attribute for humanoid soccer robots, as it directly influences their ability to make optimal decisions during gameplay. There have been several attempts to address this issue with varying success. The present study focuses on enhancing the localization accuracy in humanoid soccer robots by integrating data from Inertial Measurement Units (IMU) with camera frames and employing a Bird’s Eye View transformation to mitigate errors caused during robot navigation. In order to stabilize the camera view, the proposed method utilizes the IMU data (Pitch, Roll, Yaw) via Madgwick algorithm. The resulting Bird's Eye View frame allows precise extraction of field markers and accurate distance measurements. By computing and counteracting the robot's deviations in the x and y axes with rotation matrices, the camera frame stabilization reduces the impact of physical shocks. Experimental results show significant improvements, reducing initial positioning errors from 17.98 to 9.06 cm, an average error reduction of 8.92 cm. This novel application of Bird’s Eye View transformation advances robotic intelligence and operational efficiency in dynamic environments like RoboCup competitions.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

On a novel localization methodology for humanoid soccer robots via sensor fusion and perspective transformation

  • 期刊论文

  • 发表日期：

    2025年4月23日

查看全文
  • Farzad Nadiri,
  • Touraj Banirostam,
  • Ahmad B. Rad

摘要

Accurate localization is a fundamental attribute for humanoid soccer robots, as it directly influences their ability to make optimal decisions during gameplay. There have been several attempts to address this issue with varying success. The present study focuses on enhancing the localization accuracy in humanoid soccer robots by integrating data from Inertial Measurement Units (IMU) with camera frames and employing a Bird’s Eye View transformation to mitigate errors caused during robot navigation. In order to stabilize the camera view, the proposed method utilizes the IMU data (Pitch, Roll, Yaw) via Madgwick algorithm. The resulting Bird's Eye View frame allows precise extraction of field markers and accurate distance measurements. By computing and counteracting the robot's deviations in the x and y axes with rotation matrices, the camera frame stabilization reduces the impact of physical shocks. Experimental results show significant improvements, reducing initial positioning errors from 17.98 to 9.06 cm, an average error reduction of 8.92 cm. This novel application of Bird’s Eye View transformation advances robotic intelligence and operational efficiency in dynamic environments like RoboCup competitions.