Crane Positioning via Feature Detection

The patent introduces a marking (such as a product number or identification mark) on the plate member as an intermediary feature. Instead of relying on the physical edges or surfaces of the plate members themselves for detection, the system detects the marking position to infer the plate member's position and gravitational center. This intermediary approach enables accurate detection regardless of plate thickness, as the marking provides a consistent reference point that can be reliably captured and processed by the image acquisition and control mechanisms.

The patent replaces mechanical or manual methods for determining plate member positions with an optical detection system. The image acquisition mechanism captures images of the plate members and their markings, and the control mechanism processes these images to calculate positions and gravitational centers. This substitution of mechanical detection with optical and computational methods enables precise determination of gravitational center positions, ensuring the lifting magnet is correctly positioned over the gravitational center to prevent unbalanced loads and dropping.

The patent extracts the essential identification information from the complex visual field of stacked plate members by focusing on the markings. Instead of attempting to detect the entire stack configuration or the physical edges of multiple thin sheets, the system extracts and processes only the marking positions as key features. This extraction approach simplifies the detection problem, allowing the control mechanism to identify the uppermost sheet's gravitational center position accurately even when multiple thin sheets are stacked, thereby enabling safe and efficient lifting operations.