Edge AI, combining edge computing with artificial intelligence
The uprising of AI technology in computer vision driven by abundant sensor data,
various CNN based models, and enhanced computing power has unlocked significant potential across industries
such as ADAS/AD, drones & robotics, security & surveillance, healthcare, and more.
However, limitations of cloud-based AI such as high latency, narrow bandwidth,
and privacy risks hinder business opportunities in applications requiring real-time AI inference at the edge.
Cellplus Korea delivers comprehensive edge AI solutions by combining advanced edge computing with artificial intelligence.
We tackle the challenges of latency, bandwidth, privacy and network reliability through expert edge computing development.
Our services include AI solutions such as data annotation, image processing,
object detection, and scene analysis, enabling real-time performance where it matters most.
Immediate data processing, AI inference and video analytics at the edge, independent of network connectivity
Low size, weight and power for environments with the most limited resources and challenging conditions
Non-intrusive solutions based on data extraction, splitting, and universal interface technology
Development of customized solutions based on in-house hardware, software and firmware support
Including technologies of Super-resolution, defogging, HDR, noise reduction, image stitching, blur detection and more
Lead by experts with years of experience ensuring highest accuracy and efficiency based on know-how of utilizing minimum available data
Optimized structure design based on wide variety of options including machine learning, neural network architectures and hybrid approaches
Providing optimized AI solutions based on customer needs of accuracy, latency, purpose, environmental conditions, and budgets
· Automatic Number Plate Recognition (ANPR)
· Speed Violation Detection
· Stopped Vehicle Detection (SVD)
· Reverse Driving Detection
· AI on MCU (Micro Controller Unit)
· Edge AI Inference (Cloudless)
· Vehicle Parking Detection
· Parking Space Monitoring
· Parking Compliance Assessment
· 3D depth perception based on single NIR camera
· Forward Collision Warning (FCW)
· Object Detection in challenging weather including rain, fog, industrial smoke, sun glares and utter darkness
· Real-time performance for 30 fps
· 12 categories
· Scene Interpretation based on CNN models
· Image Captioning based on transformer architecture
· Monitoring on-site workers’ safety via recognition and captioning of safety equipment
· Image segmentation on satellite & aerial Images
· Finding cells on medical Images
· Model structure design
· AI training based on DB management
· Annotation tools including image processing functions
· Super-resolution
· Defogging
· HDR
· Noise Reduction
· Image stitching, Blur detection
Contact Us
