In order to navigate an unknown environment, an autonomous robot must be able to build a map of its surroundings while estimating its position at the same time. This problem is known as SLAM. We propose a SLAM system for stereo cameras which builds a map of objects in a scene. The system is based on the SLAM method S-PTAM and an object detection module. The object detection module uses Deep Learning to perform online detection and provide the 3d pose estimations of objects present in an input image, while S-PTAM estimates the camera pose in real time. The system was tested on a real world environment, achieving good object localization results.

中文翻译：

---

立体视觉SLAM系统上的在线目标检测和定位

为了在未知环境中导航，自主机器人必须能够构建周围环境的地图，同时估算其位置。此问题称为SLAM。我们提出了一种用于立体摄像机的SLAM系统，该系统可构建场景中的对象地图。该系统基于SLAM方法S-PTAM和对象检测模块。对象检测模块使用深度学习执行在线检测并提供输入图像中存在的对象的3d姿势估计，而S-PTAM则实时估计相机姿势。该系统在真实环境下进行了测试，获得了良好的对象定位结果。