ConductVision
Partner With Us

Capabilities - Optical Character Recognition

Optical Character Recognition reads printed and stenciled text from physical objects captured in video streams or images. The system handles ISO 6346 container codes, warehouse labels, serial numbers on equipment, safety signage, and multilingual text across 40+ languages. Unlike document-focused OCR tools, this system is purpose-built for industrial and field environments where text appears on curved surfaces, under uneven lighting, and at oblique angles. Results are structured and searchable, designed to feed directly into inventory, logistics, and compliance systems.

Optical Character Recognition
Text Extraction
Industrial
Per-Region Read Time
<200ms
Supported Languages
40+

How it works - Optical Character Recognition in three steps

Get started with optical character recognition in minutes using your existing camera infrastructure.

  • Position Cameras at Read Points. Mount cameras at container yards, conveyor belts, loading docks, or facility entrances where text needs to be captured. The system calibrates for distance, angle, and expected text regions automatically.
  • Detect and Extract Text. A text-detection model localizes text regions in each frame, then a recognition model decodes characters with language-specific dictionaries. Post-processing validates against expected formats like ISO 6346 container codes.
  • Structure and Deliver Results. Extracted text is parsed into structured fields (container number, check digit, owner code) and delivered to downstream systems via API. Misreads are flagged by confidence score for manual review.

Features

Everything you need for production-grade optical character recognition.

  • ISO 6346 Container Code Reading. Reads and validates intermodal container codes including owner code, equipment category, serial number, and check digit. Validates check digits automatically to catch misreads.
  • Serial Number Extraction. Reads serial numbers, part numbers, and asset tags from equipment, packaging, and components. Handles embossed, stamped, and printed text on metal, plastic, and cardboard surfaces.
  • Multi-Language Support. Recognizes text in 40+ languages including Latin, Cyrillic, Arabic, Chinese, Japanese, and Korean scripts. Language detection is automatic or can be specified per camera.
  • Safety Sign and Label Reading. Captures text from safety signs, hazmat placards, and regulatory labels. Useful for compliance verification and automated inventory of posted signage.
  • Curved and Angled Surface OCR. Perspective correction and dewarping algorithms handle text on cylindrical containers, angled signs, and surfaces viewed at up to 60 degrees from perpendicular.
  • Batch Image Processing. Upload folders of images for offline text extraction. Results are returned as structured JSON or CSV with image filename, detected text regions, and confidence scores.

Use cases

Real-world applications of optical character recognition across industries.

Container Yard Automation

Shipping terminals process thousands of containers daily and need accurate code reads for tracking. Automated ISO 6346 reading can replace manual data entry at gate lanes and crane operations.

  • Logistics
  • Shipping

Warehouse Inventory Verification

Warehouses need to verify that the right products are in the right locations. Camera-based label reading at dock doors and aisle intersections can catch mis-shipments before they propagate.

  • Logistics
  • Manufacturing
  • Retail

Manufacturing Quality Control

Production lines print lot numbers, date codes, and serial numbers on products. OCR can verify that printed information matches the production order, flagging discrepancies in real time.

  • Manufacturing
  • Pharma

Utility Asset Management

Utility companies maintain thousands of poles, meters, and transformers with stamped serial numbers. Mobile and drone-mounted cameras can read asset IDs during inspections, eliminating manual transcription.

  • Utilities
  • Energy

Hazmat Placard Identification

Emergency responders and facility managers need to identify hazardous materials quickly. Automated placard reading can extract UN numbers and hazard classes from truck and container signage.

  • Logistics
  • Government
  • Emergency Services

See how optical character recognition works in your environment

Schedule a free discovery call. We will walk through your cameras, your use case, and what our detection models can do for you.

Explore a Pilot

Technical specifications

Models
CRAFT text detector + TrOCR-based recognition with language-specific fine-tuning
Accuracy
Multi-language character recognition with ISO 6346 check-digit validation for container codes
Latency
<200ms per text region on NVIDIA T4
Input Formats
RTSP, ONVIF, MP4, JPEG, PNG, TIFF batch
Output Formats
JSON (structured fields), CSV, XML, Webhooks
Edge Support
NVIDIA Jetson, Intel NUC, portable ruggedized units for field use

Ready to explore what AI vision can do for you?

Start a Conversation

Contact us

  • Headquarters
    ConductScience Inc.
    5250 Old Orchard Rd Suite 300
    Skokie, IL 60077
    +1 (847) 983-3672
  • Sales & Support
    info@conductscience.com
    Mon – Fri, 9 AM – 5 PM EST