360-Degree Camera based Fast Indoor Localization using Image Gradients

在此之前，两种不同的算法在虚拟和真实环境中实现和测试。一种是基于使用线路信息，另一种是基于使用地板地图中的门信息。今年的主要任务是总结研究成果，集思广益，提高准确性。利用门信息的方法不受摄像机方向的影响，只利用图像中通过排序和归一化捕获的门的高度信息。这个概念被证明是正确的。还开发了有效的搜索方法。为了进一步提高精度和实现旋转估计，还在图像中检测了消失点。计划通过在实验装置中添加环形激光器来扩展该方法以实现3D测量。

Previously, two different algorithms were implemented and tested in virtual and real environments. One was based on using line information and the other one was based on using door information from floor maps.This year, the main tasks were to summarize the research achievements and brainstorm ideas to improve the accuracy. The method using door information is independent of camera orientation and only uses height information of doors captured in an image by ordering and normalization. The concept was proved to work properly. Efficient search methods were also developed.To further improve the accuracy and enable rotation estimation, vanishing points were also detected in images. It is planned to extend the method to enable 3D measurement by adding a ring laser to the experimental setup.