Storage-Space Image Inventorying With Automated Cargo Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Manual cargo inventorying in logistics is inefficient.
Innovation Solution
A method and device for cargo inventorying using image acquisition, inference detection, and automatic emphasis of cargo locations in storage-space images, leveraging deep learning and inference detection engines for accurate and efficient inventorying.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If manual cargo inventorying is performed, then flexibility and adaptability are maintained, but inventorying efficiency is low
Solution Approach 1:
The patent replaces the manual mechanical inventorying process with an automated computer vision system. Image acquisition devices capture storage space images, which are then processed by inference detection models to automatically identify and count cargoes. This substitution of manual mechanical operations with automated optical and computational systems directly resolves the contradiction by dramatically improving inventorying efficiency while establishing a new automation level.
Solution Approach 2:
The system enables self-service inventorying where the cargo storage system itself provides the inventory information. The image acquisition devices, inference models, and emphasis modules work autonomously to detect, identify, and highlight cargoes without human intervention. The system serves itself by using its own captured images and computational resources to generate inventory results, resolving the efficiency-automation contradiction through self-contained automated operation.
2Measurement precision
If automated inference detection is performed on storage-space images, then cargo identification accuracy is improved, but computational processing time increases
Solution Approach 1:
The system performs preliminary actions by pre-processing storage-space images to identify and emphasize cargo locations before final inventory counting. The inference detection model pre-identifies potential cargo regions, and the emphasis module pre-highlights these areas, preparing the data for faster final processing. This preliminary action reduces the computational burden during the actual inventory operation, resolving the accuracy-time contradiction by doing preparatory work in advance.
Solution Approach 2:
The patent segments the inventory process into distinct modules: image acquisition, inference detection, emphasis processing, and inventory result generation. Each module handles a specific aspect of the problem, allowing optimized processing at each stage. The inference model segments cargo identification from general image processing, while the emphasis module segments cargo highlighting from inventory counting. This segmentation enables parallel processing and optimized computational paths, reducing overall processing time while maintaining high accuracy.
Data Source
AI summary
A method and a device for cargo inventorying and a storage medium are provided in the disclosure. A storage-space image is obtained, where the storage-space image is obtained by performing image acquisition on at least one storage-space where cargoes are stored. A detection result is obtained by performing inference detection on the storage-space image. In addition, a result of cargo inventorying on the storage space is obtained by emphasizing cargoes in the storage-space image according to the detection result. In the disclosure, the detection result can be produced automatically by performing inference detection on the storage-space image. Based on the detection result produced automatically, it is also able to automatically emphasize the cargoes in the storage-space image and realize automatic inventory of cargoes by programmed automatic-control, thereby improving the efficiency of cargo inventorying.