Auto-Correct Accuracy Drift in Vision System
The accuracy drift in the vision system refers to the gradual decline in accuracy over time, typically caused by factors such as changes in the external environment and internal instability within the system.
An Auto-correct accuracy drift in vision system feature is added to the Error Analysis Tool. This chapter will introduce the application scenarios for this feature and explain how to correct the accuracy drift in EIH and ETH vision systems.
Overview
During the operation of the vision system, various factors can affect its stability and maintenance costs.
-
The significant day-night temperature variation on-site may cause fluctuations in picking accuracy, affecting the stability and accuracy of the vision system.
-
Collisions with the camera may result in changes to the extrinsic parameters. In this case, the camera should be replaced. Afterward, recalibration and resetting the pick points by jogging the robot is required, which increases workload and downtime.
-
The vision system lacks an early warning mechanism, and issues are often only detected after a major incident occurs, raising potential risks.
To effectively address these problems, the Error Analysis Tool is equipped with a built-in Auto-correct accuracy drift in vision system feature. Deploying this feature in the Error Analysis Tool can correct the accuracy drift in the vision system, improving its stability and reducing subsequent maintenance costs. The specific functionalities are described below.
-
Drift correction: It can correct drift within 10 mm in the vision system to approximately 1 mm.
Refer to the Calibration Principles section for basic principles of drift correction in the vision system.
-
Quick camera replacement: After replacing the camera, you only need to recapture the markers without any additional time-consuming operations, allowing production tasks to resume quickly and minimizing downtime.
Only cameras of the same model can be quickly replaced. If a different camera model needs to be replaced, please contact technical support for evaluation.
-
Data monitoring: The system monitors the accuracy drift of the vision system in real time and delivers early warnings when the drift compensation exceeds the alert threshold.
Application Scenarios
The “Auto-correct accuracy drift in vision system” feature applies to the following scenarios.
-
Single-camera scenarios with EIH or ETH setup
-
Multi-camera scenarios with EIH or ETH setup (Scenarios requiring point cloud stitching are not included)
|
In summary, whether the vision system uses a single camera or multiple cameras, as long as point cloud stitching is not involved, this tool can correct the accuracy drift. For vision systems with multiple cameras, the drift correction feature needs to be configured individually for each camera within the tool. Additionally, to ensure optimal drift correction results, it is highly recommended to perform drift correction on vision systems equipped with PRO S-GL, PRO M-GL, LSR L-GL, or LSR XL-GL cameras (whose working distance is 500–3000 mm). If you want to use a camera with a different working distance, please contact technical support for evaluation. |
The specific applicable and non-applicable ranges of the “Auto-correct accuracy drift in vision system” feature are shown in the table below.
| Applicable | Not applicable |
|---|---|
|
|