Dynamic OCR inspection with MVTec HALCON boosts labeling quality in regulated packaging

Strict regulatory requirements mean medical and pharmaceutical packaging must be labeled accurately before products reach customers. To meet these demands, Shenzhen Jiangrun Xuneng Technology has developed an automated OCR-based inspection system built on the standard machine vision software HALCON from MVTec. The solution significantly improves inspection efficiency and product quality for the global health and food company H&H Group.

From manual checks to automated inspection

Operating under Good Manufacturing Practice (GMP) guidelines, H&H Group must ensure that information such as batch numbers, manufacturing dates, and expiration dates is complete and correct on every carton. Previously, labeling inspections were carried out manually. High volumes and fast cycle times meant only spot checks were possible, allowing mislabeled products to reach customers—resulting in complaints, financial losses, and reputational risk.

To address this, H&H Group commissioned Jiangrun to automate the process using machine vision. The resulting platform performs 100% inspection of packaging labels, automatically reading and verifying complex alphanumeric codes at production speed.

Handling complex OCR challenges with HALCON

The application faced several challenges: encrypted and mixed-format codes, varying fonts and sizes, offset print positions, and poor print quality such as blur or incomplete characters. In addition, the system had to operate reliably at high throughput.

HALCON forms the core of the solution, handling image pre-processing, character localization, and OCR verification. The software precisely detects the character background, establishes reference lines, defines regions of interest (ROI), and compares detected characters with reference data to identify labeling errors with high accuracy.

Flexible, fast, and future-ready

“Thanks to HALCON, we significantly shortened development time, reduced overall costs, and achieved dynamic OCR with high stability and speed,” said King Wang, Technical Application Engineer at Jiangrun. “New products, fonts, or markings can be added easily, allowing the system to adapt quickly to changing requirements—even when characters are misaligned or poorly printed.”

Measurable benefits for the end user

With the automated inspection platform, H&H Group has eliminated manual label checks, reduced operator workload, and ensured consistently high labeling quality. The result is a sharp reduction in customer complaints and a stronger brand image.

Building on this success, Jiangrun plans to deploy additional machine vision applications using HALCON across other industries and use cases, supported by ongoing technical collaboration with MVTec’s local teams in China.

www.mvtec.com