AI VISION INSPECTION · PHARMACEUTICAL PACKAGING

Blister Pack
Vision Inspection System

A high-speed inline inspection system for capsule and tablet blister packs. Industrial cameras, dedicated lighting, GPU-powered deep learning, and automatic rejection work together to identify, classify, and trace defects in real time.

View Key Metrics

≥99.9%

Defect recognition accuracy

600packs/min

Maximum inspection speed

7×24

Continuous operation

3sets

Industrial cameras

Performance is subject to sample testing, production-line conditions, and final project acceptance.

PRODUCT OVERVIEW

Complex inspection,
integrated in one system

From imaging and inference to rejection and traceability, the system works with automated production lines to reduce the fatigue, variation, and data gaps of manual inspection.

Blister pack vision inspection machine
BM-AI600

Industrial imaging

Three high-resolution industrial cameras cover dedicated inspection stations.

Dedicated optics

Continuously adjustable LED lighting reveals stable, repeatable defect features.

AI inference

A high-performance industrial computer and GPU run deep learning inference in real time.

Data management

Images, parameters, batch records, and statistics are archived in one system.

High-accuracy recognition

Deep learning models and GPU inference identify a wide range of appearance and packaging defects.

High-speed inline inspection

Inspection stays synchronized with the production line to capture, classify, and reject defects in motion.

Reliable traceability

Images and inspection parameters are archived automatically for reporting and quality review.

Fast product changeover

The system adapts to different materials and batches, with sample-based model updates for new products.

INSPECTION CAPABILITIES

Detect common blister-pack defects

Inspection criteria can be configured for different materials, formats, and production batches. Images are representative defect samples.

Capsules

Inspection of fill, appearance, sealing, and foreign-material defects.

Capsules Empty cavity defect sampleEmpty cavity
Capsules Contamination defect sampleContamination
Capsules Missing cap defect sampleMissing cap
Capsules Powder leakage defect samplePowder leakage
Capsules Indentation defect sampleIndentation
Capsules Edge damage defect sampleEdge damage
Capsules Crushed blister defect sampleCrushed blister
Capsules Mesh defect defect sampleMesh defect
Capsules Print misalignment defect samplePrint misalignment
OPERATING PROCESS

Inline control,
from infeed to traceability

The machine integrates with an automated production line to perform multi-station imaging, AI classification, sorting, and data archiving in continuous motion.

  1. 01

    Automatic infeed

    Blister packs enter the inspection unit from the production line.

  2. 02

    Servo conveying

    Servo control maintains stable transport speed and inspection timing.

  3. 03

    Vacuum positioning

    A vacuum conveyor keeps each blister pack stable during inspection.

  4. 04

    Multi-station imaging

    Three industrial cameras capture images at dedicated inspection stations.

  5. 05

    AI classification

    GPU inference identifies and classifies defects in real time.

  6. 06

    Defect rejection

    The reject mechanism diverts nonconforming products to designated bins.

  7. 07

    Accepted outfeed

    Accepted products continue to the downstream packaging line.

  8. 08

    Data traceability

    Images, parameters, and batch data are archived for export and review.

TECHNICAL SPECIFICATIONS

Industrial configuration
for pharmaceutical lines

Power supply
AC 220 V ±10%, 50 Hz
Operating environment
0–45°C; 20%–80% RH, non-condensing
Compressed air
0.4–0.7 MPa, 100 L/min
Machine dimensions
2120 × 775.5 × 1811.06 mm
Machine weight
Approx. 500 kg
Vision hardware
2 MP / 5 MP industrial cameras × 3; adjustable LED lights × 3
Operator interface
15.1-inch touchscreen display
Data capabilities
Image and parameter archiving, Excel export, project-based MES integration
Blister inspection machine dimension drawing
Dimensions are indicative. Final project drawings take precedence.
ABOUT ZHONGHENG VISION

Machine vision for
more reliable quality decisions

Zhongheng Vision Intelligence Technology (Hangzhou) Co., Ltd. is an innovative technology company focused on artificial intelligence, deep learning, and machine vision algorithms, integrating the development, production, sales, and service of visual inspection equipment.

Our team develops AOI, AI, laser, and 3D imaging technologies for pharmaceutical, packaging, semiconductor, and magnetic-material applications, helping manufacturers stabilize quality, improve efficiency, and reduce manual inspection costs.

PROJECT CONSULTATION

Bring your samples and line data.
Let us validate the solution.

We evaluate the product type, defect criteria, production speed, and available floor space to define the right inspection configuration.

PhonePhone number pending
AddressCompany address pending
WeChat QR code pending
CONTACT FORM

Tell us what you need

Leave any one of the following fields and our team will follow up.