Qingfeng Li Email & Phone Number

北京

杭州

 Project: loop-closure based on visual inertial Describe: Our loop-closure only optimize 4-dof pose owe to pitch and roll angles are directly obtained by VIO. We use FAST keypoints and BRIEF descriptors. In order to detect revisited places we use an image database composed of a hierarchical bag of words. The bag of words is a technique that uses a visual.

Singapore

 Project: Real-time Vehicle Ego-Motion Estimation Describe: Our group take charge stereo visual odometry. At present, we have a difficult on estimating the scale of features in large scale and high-speed pose change environment. We present a method that careful select features and tracking to improve the accuracy of feature detect and match. What’s more.

深圳

 Project: A wide range of 3d reconstruction based on monocular visionDescribe: Our group utilized UVA to shoot videos in the large scale environment of outdoor.Then we reconstructed the scene of videos by using improved 3D reconstruction method. Firstly, we used monocular odometry of libviso2 to approximately estimate initial iterative pose. Secondly, we.

硕士, Computer

Master'S Degree, Control Science And Engineer, Good

Activities and Societies: NO Paper: Robust Monocular Visual Inertial using nonlinear optimization(2017 ICIA) Paper: Mobile Robot.

Bachelor'S Degree, Automation, 85

Activities and Societies: Arts department  Competition: The Freescale “Cup” Intelligent Car Competition Primary duty：My main tasks.