Depth Sensor Shelf Stock Estimation

The patent replaces the manual mechanical checking process with an automated optical system. A depth sensor (time-of-flight camera) captures 3D images of shelf contents, and a processing system automatically analyzes the depth data to detect product presence, estimate stock levels, and identify price tag locations. This substitution of mechanical manual inspection with optical sensing and computational analysis simultaneously improves accuracy (through precise depth measurement) and reduces time consumption (through automated parallel processing of multiple shelf locations).

The system enables self-service inventory monitoring by allowing the shelf scanning system to automatically detect and report stock levels without human intervention. The depth sensor autonomously captures 3D data, the processing system automatically analyzes the data to determine product presence and quantity, and the system self-updates inventory records. This automated self-service approach eliminates the need for manual checking while maintaining high accuracy and significantly reducing time consumption.

The patent replaces complex manual inventory management processes with a streamlined automated system using time-of-flight depth sensors and image processing algorithms. The depth sensor captures 3D spatial information in a single shot, eliminating the need for multiple manual measurements or complex mechanical scanning mechanisms. The processing system uses computational algorithms to automatically interpret depth data, determine product presence, and calculate stock levels, thereby achieving high productivity through software-based automation rather than complex mechanical systems.

The system achieves high productivity by changing the measurement parameter from 2D image capture to 3D depth measurement using time-of-flight sensing. This parameter change allows the system to obtain spatial information about product location and quantity in a single capture event, significantly improving scanning speed. The depth data provides direct information about product presence and stock levels without requiring complex mechanical disassembly or multiple measurement steps, thus enhancing productivity while keeping the hardware relatively simple.

The patent replaces energy-intensive continuous mechanical scanning with a single-shot time-of-flight depth sensing approach. The depth sensor captures the entire shelf scene in one instantaneous 3D measurement, eliminating the need for continuous scanning or multiple sequential measurements. This substitution reduces computational energy requirements because the system processes a single depth map rather than continuously analyzing multiple frames or performing repeated mechanical scans, while still achieving high stock level estimation accuracy through the rich spatial information contained in the 3D depth data.

The system applies partial action by selectively analyzing only the relevant portions of the depth data corresponding to product locations, rather than processing the entire 3D scene in full detail. The processing system identifies regions containing products and focuses computational resources on those specific areas for stock level estimation. This selective processing approach reduces overall computational energy consumption while maintaining high accuracy for the critical measurement of product inventory levels.