Abstract: In low-texture scenes, Visual Odometry (VO) algorithms often encounter challenges stemming from sparse feature sets and reduced accuracy in feature matching. To overcome this, integrating ...