6️⃣3D Visual Positioning Soft Package De-Stacking Solution Introduction

Industry challenges of soft package de-stacking

• Soft package deformation: Soft package is easy to deform during the stacking process, resulting in increased difficulty in grasping.

• Low efficiency: Manual dismantling takes a long time, and long-term high-intensity operation is prone to errors.

• Safety hazards: Leakage and breakage under traditional operation not only increase additional material loss but also pose safety risks.

Technical highlights of MRDVS's soft package de-stacking solution

• High-performance RGB-D camera and AI algorithm
The camera can capture surface texture information and point cloud data of the stack at the same time. Combined with deep learning of AI algorithms, the system can accurately identify and locate irregular objects such as soft packages and cardboard boxes, fully utilizing their surface texture and edge features to ensure the stability and accuracy of each capture action, effectively reducing packet dropping and breaking phenomena.

• Algorithm integration on the camera side
The stacking algorithm runs inside the camera, achieving real-time data processing and reducing dependence on external industrial control computers. This design greatly simplifies system deployment, reduces overall costs, and improves system response speed and operational efficiency.

• Strong anti-light ability
The system has excellent anti-light ability and can maintain stable visual recognition performance in complex lighting environments. Whether it is strong light or weak light conditions, the system can effectively capture the details of objects, ensuring the smooth execution of de-stacking tasks.

• Self-Adaptation Recognition
There is no need to collect training data in advance. The system can self-adapt to recognize the shape of soft packages, supporting complex structures such as six-pattern stacks, five-pattern stacks, and spiral stacks, achieving efficient de-stacking and reducing debugging difficulty.

• Sequential planning based on soft package position
By obtaining the location information of each soft package, a reasonable grabbing sequence planning is provided. Even in the face of complex stacking forms, the system can ensure that each soft package is efficiently and safely processed.

• Four-point calibration, simple and efficient
MRDVS provides a four-point calibration method, allowing users to flexibly choose calibration points. The system automatically calculates the conversion matrix between the camera and the robotic arm, which is more convenient than traditional 9-point or 64-point calibration.

Wide adaptability and integration capabilities

Economic value and application prospects