Filling Line Vision Inspection
Pre-Filling Inspection
A well-known beverage enterprise has introduced Newvin visual inspection system to strictly control the appearance of empty glass bottles.On the production line, multiple sets of high-precision industrial cameras rapidly capture images of empty bottles, covering the bottle mouths, bodies, and bottoms.
Equipped with customized ring and backlight sources, the system effectively eliminates reflections and highlights defects.
Leveraging advanced image recognition algorithms, the inspection system accurately identifies flaws such as bottle mouth notches, body scratches, bottom cracks, and residual impurities inside the bottles.
It boasts an inspection speed of 12,000 bottles per hour with an accuracy rate of over 99%. Once a defective bottle is detected, the automatic rejection device is activated immediately to divert it from the production line.Since the system was put into use, the defective product rate has been reduced by 80%, significantly improving product quality and production efficiency.
Surface defects - Scratches, bubbles, inclusions, dirt stains, oil residues, mold release agent deposits.
Structural defects - Cracked necks, chipped rims, incomplete threads, deformed bases, flash residues.
100% Inspection: Eliminates manual sampling errors and ensures compliance with safety regulations for food and pharmaceuticals.
High Speed and Low Latency: Matches the pace of high-volume manufacturing lines without bottlenecks.
Reduced Scrap: Catches defects early in production, minimizing waste from downstream filling and packaging processes.
Predictive Maintenance: Monitors mold wear over time by tracking dimensional deviations, enabling proactive mold replacement before defects occur.
Post-Filling Inspection
This is a critical quality control process widely adopted in industries such as pharmaceuticals, food and beverage, cosmetics, and chemical manufacturing. It focuses on verifying product integrity and compliance after the filling and sealing stage of production.
Checks for correct liquid fill levels(Volume).
Cap sealing inspection
Misaligned & Missing Cap Detection.
End-of-Line Verification. （Label and Printing Defects）
Misaligned labels, smudged printing, missing barcodes, incorrect expiration dates.
Industry Applications
Food & Beverage: Inspects beer, wine, and soda bottles for contamination and labeling accuracy
Pharmaceuticals: Ensures sterile injection vials and IV containers meet USP and cGMP standards
Cosmetics: Verifies premium glass packaging aesthetics
Chemicals: Detects structural defects in industrial chemical containers.
Core Components of Newvin Vision Inspection Systems
High-Resolution Cameras: Capture detailed images of the bottle surface, enabling the detection of even the smallest defects.
Advanced Lighting: Utilizes specialized lighting techniques (e.g., backlighting, diffuse lighting) to enhance defect visibility and contrast.
Powerful Image Processing Software: Newvin's proprietary software analyzes images in real-time, applying algorithms to detect and classify defects based on predefined criteria.
Flexible Integration: Systems can be integrated into existing production lines, with options for standalone or inline inspection setups.
How Newvin Vision Inspection Works
Image Acquisition: High-resolution cameras capture images of the glass bottle as it moves along the production line.
Preprocessing: Images are preprocessed to enhance contrast, reduce noise, and correct for lighting variations.
Defect Detection: Advanced algorithms analyze the preprocessed images to identify potential defects. This may involve pattern recognition, edge detection, or texture analysis.
Defect Classification: Detected defects are classified based on their type, severity, and location. This allows for prioritized sorting and rejection of defective bottles.
Data Logging and Reporting: Inspection results are logged for traceability and quality control purposes. Reports can be generated to provide insights into defect trends and production line performance.
Technical Principles
Image Acquisition
Industrial cameras capture images of the filling area, covering key components such as product appearance, labels, and seals.
Preprocessing
Operations such as denoising, contrast enhancement, and lighting adjustment are performed to improve detection accuracy.
Feature Extraction
Algorithms identify parameters such as fill volume, position, and shape, and compare them against predefined standard
Defect Classification
Deep learning models (e.g., CNNs) classify compliant and defective products (e.g., underfilling, overflow, misaligned labels).
Result Output
Detection reports are generated to trigger sorting or alarm systems, enabling automated quality control.