Skip to content
TH
📦 Container OCR

Automatic Container
Code Recognition

AI system for automatic container number reading per ISO 6346 standard for port terminals and logistics operations.

Automatic ContainerCode Recognition

Usage Context and Data Preparation

Container codes contain an owner code, equipment category, serial number and check digit as described in the existing features. Reading the code and validating its check digit address different checks. A valid format does not necessarily mean the container matches the intended shipment.

Prepare images from the actual crane, gate or forklift camera positions, together with sample codes and TOS/WMS reference records. Define what counts as one transaction and how images, times and positions are reviewed when searching for a record or correcting an incomplete code.

ISO 6346 describes container-code structure and supports format checking; mentioning it does not certify this product. Preserve the complete required characters and distinguish the identifier from surrounding text such as weights and other markings on the container wall.

Check digits detect some error patterns, not every OCR error or a match to shipment records. Compare reference records and actual events. Failed or incomplete codes need review; do not alter characters merely to satisfy a checksum without image evidence.

Site images may contain rust, paint damage, vertical text and structural obstructions. Select views from the actual handling sequence. Multiple views must be associated with the same container, not neighboring containers or separate transactions occurring close in time.

TOS/WMS integration needs transaction identity, reading point, time and review ownership. One container appears at several workflow stages, so its code alone cannot deduplicate events. Human corrections should preserve original readings and reasons within the agreed traceability scope.

Evaluate complete-code accuracy and transaction association. Distinguish incomplete visibility from recognition errors, and group results by angle, character condition and layout. Valid format alone is insufficient because a correct-looking code can still be attached to the wrong transaction.

Multiple-side reading, shipment verification and evidence-review workflows require proposal confirmation and trials with actual systems. Container-code reading does not establish internal contents, seal integrity or customs status; those conclusions need separate evidence and responsible procedures.

✨ Key Features

Key Features

📝

ISO 6346 Compliant

Read complete container codes: Owner Code, Equipment Category, Serial Number, and Check Digit.

📸

Auto Detection

Detect containers from cameras mounted on cranes, gates, or forklifts.

✅

Check Digit Validation

Automatic check digit verification to eliminate errors.

🔗

TOS/WMS Integration

Direct integration with Terminal Operating Systems and Warehouse Management Systems.

📊

Transaction History

Log every transaction with photos, timestamps, and positions for historical search.

🌧️

All-Weather Operation

Operates in all conditions — day, night, rain, and bright sunlight.

Scroll the diagram horizontally to inspect all details.

Container Code Recognition System
Container Code Recognition SystemConceptual illustration, not a product screenshot or measured result.
🎯 Key Benefits

Key Benefits

⏱️

90% Time Reduction

From 30 seconds manual entry to 1 second automatic.

❌

Error Elimination

No more typos or miskeyed data — prevents container mix-ups.

📈

Increased Throughput

Handle growing container volumes without adding staff.

🔄

Real-time Tracking

Track containers from gate-in to gate-out instantly.

⚙️ Technical Specifications

Technical Specifications

Item Specification
Accuracy > 98%
Read Speed < 1 second
Standard ISO 6346
Max Vehicle Speed 30 km/h
Supported Cameras IP Camera 5MP+
Integration REST API, TCP/IP, MQTT
🏭 Suitable For

Suitable For

Port Terminals
Container Yards
Weighbridge / Gate Access
Distribution Centers
Customs Operations

Related articles and knowledge

Vision AI Knowledge Library

Start with image interpretation, event information and evidence-backed retrieval.

Read More

Detection and Event Indexing

Define identifiers, time, attributes and data lifecycle before choosing an index.

Read More

Natural-Language Search

Separate objects, attributes, relationships and questions requiring confirmation.

Read More

Hybrid and Multimodal Retrieval

Combine filters and vectors without assuming similarity proves every condition.

Read More

Search Evaluation and Governance

Evaluate relevance, evidence, permissions and information missing from an index.

Read More

Interested in AI Vision for Your Organization?

Contact us for a free consultation and live system demo.

Explore AI Technology and Practical Guides

Read published articles on principles, limitations and system evaluation.

News