V-ONE’s Solution
-
Medical & Pharmaceutical
- Ultrasonic imaging diagnosis
- Reagent classification
- Culture cell inspection
- Syringe inspection
- Fluid foreign body inspection
- Pill surface inspection
-
Display & Secondary Battery
- Indentation foreign body inspection
- Inspection of foreign substances, cracks and burrs
- Scratch inspection
- Cell phone surface inspection
- Lead defect inspection
- Stain inspection
-
Spare Parts & Material
- Metal bobbin surface inspection
- Liquid capsule contents inspection
- Rubber material surface inspection
- Wire classification/defect inspection
- Lead application inspection
- Wafer crack inspection
-
Logistics & Distribution
- Box/Packaging classification
- Assemble parts classification
- Monitoring for loss of items
- Agricultural goods automatic classification
- Consumer orders inspection
- Packaging/Container OCR reading
withAI
Quickly develop solutions by using various models based on the V-ONE Solution*, and work easily, clearly through a user-friendly interface.
-
PRE Block
Pre-processing algorithm to improve detection power of inspection images
-
AI Block
API for learning and inference (inspection)
-
APP Block
API for AI project
V-ONE M/V solution has been recognized for its technological prowess and quality excellence in successfully carrying out various machine vision projects such as displays and secondary batteries.
Feature
-
V-ONE solution has a scalability and flexibility respond to various business needs.
-
Provides the ability to inspect various models in one task as needed.
-
Items that could not be inspected using existing inspection equipment can be detected.
-
Rapid inspection possible by supporting self-developed high-speed processing technology and multi-parallel processing technology.
-
Supports an optimization module for inspecting large amounts of image data.
-
Based on the block structure, users can only use the blocks they want.
-
Detection power is improved by suggesting specialized models depending on the inspection target.
-
Supports models that can be used in various fields such as manufacturing, medical care, materials, parts, distribution, and logistics.
Constitution
-
-
Customer
request -
Data collect & processing
-
-
PRE Block
- Customer data processing
- studying by Processed image
- rule base pre processing algorithm for increase inspection power
-
AI Block
Object
DetectionSegmentation
Classification
Anomaly
- Easy and intuitive work with labeling tools provided
- Detection class definition
- Start learning immediately after checking labeling
- Continuous model advancement
- Multi-learning support (Multi GPU) model creation
- Optimization model transformation
- Model validation
- Model quality check by appropriate inspector
-
-
APP Block
-
V-ONE
Solution
- Inspection model provided
- API support for AI projects
- Supports S/W customization in the way the customer needs
- Support method (project, module, consulting)
- Equipment development*
- Optical module + model + API*
- Model + API*
-
-
-
Support model & Consulting, Project
-
Customer
-
-
Equipment*
-
- AOI
- AMI
- DMS
- Etc.
-
withAI
-
-
Part’s (Module)*
-
- Optic
- V-one Solution
- withAI
-
- Display
- Secondary battery
- Medical
- Material
-
-
Application*
-
withAI
-
API
-
Detail View
withAI
PRE BlockPreprocess
- Auto-Labeling
- Auto-Tuning
- GAN (Generative Adversarial Networks)
- Algorithm for improving video quality
- Algorithm which is adjust shape for suitable image of model
- Algorithm of simplifying model to improve model accuracy
- Algorithm for classification
- Algorithm of converting frequency for speckle inspection
- Algorithm of feature extraction from image
- Algorithm of finding shape in image information
- Neural network (MLP)
- Customizing Module
withAI
APP BlockApplication
- Provide public Lib for AI model inspection
- ONNX model provided
- Optimized engine providing for inspection time reducing
- C/S App through learning and inspection server operation provided
- Mobile model provided
- Provides for apps tailored to customer needs
- Industry-specific business models provided
- Customizing Module
withAI
AI BlockObject Detection
- After classifying the object and detecting its location If it is determined to be an object, it is displayed as a rectangle
- Detect multiple objects in one image
- Using pre- trained model, improve detection accuracy and time reduction
- Select an appropriate model according to the process and task pre-learning and inspection
Medical / Pharmaceutical
- Model dedicated to pill surface inspection provided
- Model dedicated to the inspection of foreign substances in injected fluid. Provided
- Model dedicated to specimen bin classification provided
Materials
- Model dedicated to metallic appearance inspection provided
- Model dedicated to surface inspection of rubber materials provided
- Model dedicated to electronic component/board inspection provided
- Model dedicated to wafer surface inspection provided
Logistics
- Model dedicated to product classification provided
- Model dedicated to classification of agricultural products provided
- OCR reading model provided
Display
- Model dedicated to edge crack inspection provided
- model dedicated to lead burr inspection is provided
- Model exclusively for mobile phone surface inspection is provided
Segmentation
- Classify objects by dividing into pixels, display location
- Detect multiple objects in one image
- Using pre- trained model, improve detection accuracy and time reduction
- Select an appropriate model according to the process and task pre-learning and inspection
Med / Pharmaceutical
- Ultrasound imaging diagnostic model provided
- Cultured cell inspection & classification models provided
Materials
- Model dedicated to wafer surface inspection provided
Display
- Speckle inspection model provided
- inspection models for cracks, foreign substances, contamination provided
Classification
- Determine object type and classified
- Determine one object in one image
- Using pre- trained model, improve detection accuracy and time reduction
- Expected improvement in output when fused with Pre-Block
- Select an appropriate model according to the process and task pre-learning and inspection
Med / Pharmaceutical
- Drug classification model provided
Logistics
- Pre-Block combination, OCR reading model provided
Anomaly
- Judgment into normal and abnormal binary types
- Judged as normal/bad in one video
- Learning time is fast as it learns only from normal images
- Collecting poor training data Improves training time and detection accuracy by using pre-trained models
Parts/Materials
- Provides a model for inspecting the presence or absence of lead coating on the board
Logistics/Distribution
- Box damage inspection model provided
Applying Field
-
Medical & Pharmaceutical
-
Ultrasonic imaging diagnosis
-
Culture cell test - Live, Dead, Debris classification
-
Medicinal type/grade classification
-
Foreign substances in injected fluid, print inspection.
-
Classification by specimen container type
-
Pill breaks, scratches, color, and shape inspection
-
-
Material & Parts
-
Inspection of metal bobbins for nicks and scratches
-
Inspection of inner cable color for wire cross sections
-
Inspection of capsule contents for bubbles and bursting
-
Inspection of wafer scratch and speckle
-
Inspection of lead soldering application (none lead, over lead, less lead, short)
-
Inspection of rubber ring colored foreign matter, tear, and edge
-
-
Display & Secondary battery (Rechargeable)
-
Inspection for Edge crack & damage
-
Lead burr, open, short defect inspection
-
Speckle (dot, line, north pole) inspection
-
Mobile phone scratches, camera and fingerprint sensor scratches, and surface inspection for cracks
-
-
Logistics & Distribution
-
Sorting boxes on the move
-
sorting screws which for assembly of finished products
-
Automatic classification of agricultural products by size
-
Inspection of customer order item & ship and return product
-
OCR reading of packaging, containers, etc.
-
-
Foods & Packaging
-
Inspection of cosmetic containers for breaks, scratches, and foreign substances
-
Inspection of expiration date printing on packaging
-
Inspection of Can, bottle, canned food printing and damage
-
Inspection of contamination and foreign substances in food
-
-
Robotics Autonomous Driving
-
AMR (Autonomous Mobile Robots)
-
Location measurement and real time map generation
-
Path creation, avoiding object, collision prevention through Object detection technology
-
Prediction of drivable area using Segmentation
-
Reinforcement Deep learning-based multi robot and avoid obstacle, abnormal detection
-
Real time collaborative detection with camera detection and DDS combined sensors
-
Product List
withAILite |
withAIProfessional |
withAIMedical |
|||
---|---|---|---|---|---|
Object-Detection | |||||
Instance Segmentation | |||||
Classification | |||||
Anomaly | |||||
Semantics Segmentation | |||||
Learning Model | Small | Large | Large | ||
Multi GPU | 1 | 2 | 2 | ||
ONNX | |||||
Optimized Inference | |||||
Multi Threading | |||||
Auto Learning | |||||
Customizing | |||||
Logistics / Distribution |
Industrial |
Medical / Pharmaceutical |
Specification
withAI Lite
withAI Professional
division | Minimum specifications | Recommended specifications | |
---|---|---|---|
Train PC | CPU | Intel i7 or more | Intel i7 or more |
RAM | 16GB | 32GB | |
CUDA Compute Capability | 3.5 | 3.5 | |
GPU | Geforce RTX 4070 (12GB or more) | Geforce RTX 4090 (24GB or more) | |
OS | Windows 10 x64, Windows 11 x64 | ||
Inference PC | CPU | Intel i7 or more | Intel i7 or more |
RAM | 16GB or more | 32GB or more | |
GPU | Geforce RTX 4070 (12GB or more) | Geforce RTX 4080 (16GB or more) | |
OS | Windows 10 x64, Windows 11 x64 | ||
Develop environment | Visual Studio 2017 or more |
withAI Medical
division | Minimum specifications | Recommended specifications | |
---|---|---|---|
Train PC | CPU | Intel i7 or more | |
RAM | 32GB | ||
CUDA Compute Capability | 3.5 | ||
GPU | GeForce RTX 4090 (24GB or more) | ||
OS | Windows 10 x64, Windows 11 x64 | ||
Inference PC | CPU | Intel i7 or more | Intel i7 or more |
RAM | 16GB or more | 32GB or more | |
GPU | Geforce RTX 4070 (12GB or more) | Geforce RTX 4080 (16GB or more) | |
OS | Windows 10 x64, Windows 11 x64 | ||
Develop environment | Visual Studio 2017 or more |