Handle complex defects, weak features, small samples & rule-algorithm limits
Covers AI Platform, Inspection SW, Host System & Vision Device.
Connect inspection data, equipment status, logs, maintenance & BI for quality traceability, equipment health and workshop self closed-loop management.
Support sample validation, design, joint commissioning & on-site rollout.
Import industrial images for training or inspection. The system automatically validates format and integrity.
Build serial or parallel vision processing pipelines.
Combines manual & AI-assisted annotation to build industrial vision datasets for detection, segmentation, classification and anomaly analysis.
Expand lighting, scale and pose distribution of industrial images, and standardly split datasets for training, validation and test.
Configure tasks, models and computing resources, and monitor training loss, validation metrics and weight saving status in real time.
Evaluate model convergence and verify segmentation by single & batch inference.
Convert trained weights to ONNX, packing input size, classes and normalization for industrial deployment.
Centralized view of projects, equipment status, operation data and recent events.
Create, open and maintain vision inspection solutions, displaying workflows and device configurations.
Build inspection workflows with modular nodes using existing VM-0522 plugins.
Unified management of industrial cameras, motion control cards, axis parameters and device connection status.
Configure communications for PLC, TCP, serial port, WebService, light sources and external instruments.
Centralized view and maintenance of shared data among system, solutions, equipment and workflow modules.
Analyze inspection yield, cycle time, defect distribution and track system events.
Configure UI, runtime parameters, permissions, canvas and data storage policies.