Conveyor Motion Estimation Using Patch Displacement in Homogeneous Scenes

The patent segments the image into multiple patches and processes each patch independently to estimate displacement. This segmentation approach allows the system to handle homogenous conveyor environments by focusing on local motion patterns in each patch rather than requiring global scene heterogeneity. The segmentation principle directly addresses the limitation of existing techniques that rely on scene-wide motion cues.

The patent applies partial action by estimating displacement for only those patches that contain moving objects or conveyors, rather than requiring all patches to contribute to motion estimation. This allows the system to work effectively in scenarios where motion is partially occluded or confined to specific regions of the image, directly addressing the limitation of existing techniques that require comprehensive scene motion.

The patent implements self-service by enabling the robotic system to autonomously estimate conveyor motion using its own sensor suite without requiring communication with external conveyor controllers. The system processes images from its cameras, identifies patches containing conveyors or objects, and calculates displacement independently. This self-service capability directly improves reconfigurability while maintaining reliability through robust image-based motion estimation.

The patent replaces mechanical or electrical coupling with conveyor controllers by using vision-based sensing and image processing to estimate motion. Instead of relying on mechanical sensors or communication protocols, the system substitutes these with optical sensing and computational methods to derive conveyor speed and position, thereby improving system independence and reconfigurability.

The patent divides the image into multiple patches and processes each patch independently to estimate displacement. This segmentation reduces computational complexity by allowing parallel processing of smaller regions rather than analyzing the entire image as a single unit. The segmentation enables continuous motion estimation while managing computational load through distributed processing.

The patent applies partial action by focusing computational resources only on patches that contain conveyors or moving objects, rather than processing every patch uniformly. This selective approach reduces overall computational complexity while maintaining continuous motion estimation accuracy by concentrating effort on relevant regions of the image.