Visualization software suite SMore ViMo

image-processingdata managementmodeling

Product Description SMore ViMo series products are equipped with intelligent production models for over 1000 segmented industry scenarios, embedded with over 50 tools such as algorithm enhancement, and provide full-stack intelligent capabilities for the entire manufacturing process. SMore ViMo covers industrial intelligent clouds, deep learning training software, machine vision software and other industrial products, delivering data management, visual solution design, and end-to-end runtime software to meet industrial visual needs. The platform enables factories to quickly implement model training, software integration and production line deployment without relying on professional algorithm personnel, lowering the threshold for intelligent applications in manufacturing and providing inclusive intelligence for industry.

Product Features

• Data management and visual solution design for industrial applications.
• End-to-end running software enabling model training, software integration and deployment.
• Enables non-algorithm and non-vision engineers to build algorithm modules and train high-precision models via drag-and-click (zero-code / low-code).
• Product matrix includes ViMo Cloud, ViMo Deeplearning and ViMo Studio.

Major Advantages

• High Precision Algorithm: Built with multiple deep learning algorithms and hundreds of traditional operators; tailored high-precision visual algorithms for manufacturing with strong performance in multiple industry projects.
• Zero code / Low code: Universal platform ready to use out of the box; allows rapid construction of algorithm modules and model training through graphical operations.
• Multi-scenario Universal: Supports over 1000 segmented application scenarios across industries such as new energy, pan-semiconductor, automotive, consumer electronics, food and healthcare; flexible for complex industrial visual inspection solution designs.

Application Cases

• USB Defect Detection for a Consumer Electronics Device Manufacturer: 3-class classification (OK/NG/NG-2) for scratches and dirt on USB interface. Transitioned from manual inspection to automated deep-learning detection achieving accuracy and recall around 98.9% with false negative rate below 2%, enabling full automation and improved yield.
• Defect Detection for Small-sized Lithium-ion Battery Casing: Replaced manual visual inspection with SMore ViMo and Vi-Studio design, enabling detection of leakage glue, scatter points and broken glue; achieved fast cycle times (photo-calculation-material dispensing within 500ms per piece) and facilitated quick line deployment.
• Vision Inspection of Silicon Wafer Characters: Equipped with OCR, defect segmentation and detection modules and a 100-million-pixel resolution algorithm; supports dynamic data augmentation and few-shot learning to enable intelligent full inspection of high-resolution wafer images.

Additional Application Highlights / Case Studies

• OCR for Smartwatches: Character recognition on metal components under challenging imaging conditions (blurriness, glare, tilt) achieving recognition accuracy up to 99.9% and deployment across multiple automated production lines.
• OCR Detection for Vibrating Motor Magnet Blocks: Handles extremely small and reflective engraved characters (down to 0.1 size) with robust pre-processing and recognition performance.
• Wireless Charging Coil Inspection for Earphones: Multi-class detection covering 20+ defect types with pixel-level processing and throughput enabling 100% production coverage with a fraction of manual inspectors; processing speed cited ~0.2s per piece.
• Appearance Inspection of Bearings: Integrated optical solution (multi-light sources, edge-compute machine, color and 4K line-scan cameras) with ViMo algorithms to recognize 20+ defect types (rust, oil stains, scratches, etc.) and achieve high detection precision for automotive bearing inspection.
• Diode Defect Detection: Segmentation algorithms detecting >10 defect types (fractures, bubbles, shallow edge cracks) with segmentation accuracy surpassing traditional methods and manual inspection, achieving low false positive/negative rates.

Product Matrix & Ecosystem

• ViMo Cloud: industrial intelligent cloud capabilities for data and deployment orchestration.
• ViMo Deeplearning: deep learning training software and high-precision algorithm toolset.
• ViMo Studio: visual solution design and zero/low-code algorithm composition and deployment tools.

Caractéristiques / spécifications techniques

• Supports >1000 segmented industry scenarios.
• Contains >50 embedded tools for algorithm enhancement and solution design.
• High-precision algorithms combining multiple deep learning methods and 100+ traditional operators.
• Zero-code / low-code graphical design and model training workflow.
• Capabilities: data management, visual solution design, end-to-end runtime software and operation management (Vi-Lab) for rapid delivery.
• High-resolution image processing: examples include 100-million-pixel resolution algorithm for large images.
• Advanced modules: OCR, defect segmentation, multi-class classification, pixel-level detection and semantic matching.
• Supporting hardware integrations: edge computing integrated machines, color cameras, 4K color line-scan cameras and multi-light optical solutions.
• Typical achieved metrics in case studies: detection accuracy/recall up to ~98.9%, recognition accuracy up to 99.9%, processing speeds cited 0.2–500 ms per piece depending on scenario.