Wide range of AI visual inspection applications

Solvision has delivered outstanding results in the semiconductor, LCD display, footwear, textile, automotive, welding, and a wide range of other industries. Its flexibility has allowed it to be implemented in several applications such as defect inspection, presence detection, counting, classification, optical character recognition, and many more.

Wide range of AI visual inspection applications

Few training samples for deep learning required

Solvision offers many easy-to-use data augmentation tools that allow users to simulate different real-life scenarios. Needing only 1/10 of the samples typically required by AI inspection software, our vision solution significantly reduces the amount of time that engineers have to spend during the labeling process.

User-friendly vision system interface

Our user-friendly design allows users to label several defect types at the same time, a very convenient feature in applications where multiple defects and features need to be simultaneously classified.

SOLOMON 3D - JustPick智能分揀系統 -採開放式平台

Simple industrial robot and PLC integration

Solvision provides easy integration with more than 20 robot brands and built-in PLC communications through the TCP/IP and Modbus communications protocol at no extra cost, allowing users and system integrators to choose the product that they feel most comfortable with.

Parallel defect detections

Solvision can select multiple GPUs and graphic cards to disperse the AI computing load, allowing the users to carry out simultaneous detections at the same time.

Wide range of AI visual inspection applications

Solvision has delivered outstanding results in the semiconductor, LCD display, footwear, textile, automotive, welding, and a wide range of other industries. Its flexibility has allowed it to be implemented in several applications such as defect inspection, presence detection, counting, classification, optical character recognition, and many more.

Few training samples for deep learning required

Solvision offers many easy-to-use data augmentation tools that allow users to simulate different real-life scenarios. Needing only 1/10 of the samples typically required by AI inspection software, our vision solution significantly reduces the amount of time that engineers have to spend during the labeling process.

User-friendly vision system interface

Our user-friendly design allows users to label several defect types at the same time, a very convenient feature in applications where multiple defects and features need to be simultaneously classified.

SOLOMON 3D - JustPick智能分揀系統 -採開放式平台

Simple industrial robot and PLC integration

Solvision provides easy integration with more than 20 robot brands and built-in PLC communications through the TCP/IP and Modbus communications protocol at no extra cost, allowing users and system integrators to choose the product that they feel most comfortable with.

Parallel defect detections

Solvision can select multiple GPUs and graphic cards to disperse the AI computing load, allowing the users to carry out simultaneous detections at the same time.

Solvision  Applications

Presence/Absence Detection

Push-through-package (PTP) blister production lines on average pack 5,000-40,000 tablets or capsules per hour, and are prone to occasional errors in filling. These production errors can range from unfilled and deformed blisters to inadequate or broken tablets or capsules.

Defect Identification

Mildew and stains sometimes remain in the bottle even after the disinfection process. However, checking for defects involves rotating and moving the bottle, and there is usually a product label in the way that makes it unfavorable for manual or traditional vision systems.

Categorization

Based on deep learning technology, Solvision can locate and mark the position of eggshell defects on sample images to train an AI inspection system. The AI model can then detect pores and cracks on the eggshell surface and classify eggs into the trained categories to meet safety standards and increase commodity value.

Solvision production line defect inspection 

Metal scratch defects

金屬瑕疵檢測

Chicken nugget defects

冷凍雞塊瑕疵檢測

Contact lenses defects

隱形眼鏡瑕疵檢測

Laser welding defects

雷射焊接瑕疵檢測

Solvision  Applications

Presence/

Absence Detection

Push-through-package (PTP) blister production lines on average pack 5,000-40,000 tablets or capsules per hour, and are prone to occasional errors in filling. These production errors can range from unfilled and deformed blisters to inadequate or broken tablets or capsules.

Defect Identification

Mildew and stains sometimes remain in the bottle even after the disinfection process. However, checking for defects involves rotating and moving the bottle, and there is usually a product label in the way that makes it unfavorable for manual or traditional vision systems.

Categorization

Based on deep learning technology, Solvision can locate and mark the position of eggshell defects on sample images to train an AI inspection system. The AI model can then detect pores and cracks on the eggshell surface and classify eggs into the trained categories to meet safety standards and increase commodity value.

Solvision production line defect inspection 

Metal scratch defects

金屬瑕疵檢測

Chicken nugget defects

冷凍雞塊瑕疵檢測

Contact lenses defects

隱形眼鏡瑕疵檢測

Laser welding defects

雷射焊接瑕疵檢測

Specifications

Module NameSLM VISAI-S100
Operating SystemWindows 10 (64 bit)
CPUMinimum: Intel Core i5 / Recommended: Intel Core i7
GPUMinimum: Nvidia GTX 1070 (RAM: 8GB) / Recommended: Nvidia RTX2080 Ti
RAMMinimum: 16G / Recommended: 32G
InterfaceCamera Interface
Coding InterfaceNet framework 4.8
Coding LanguageC #
Compatible WithMVTec Halcon, NI LabVIEW
LanguageEnglish
Image FormatPNG, BMP, JPG, JPEG, JPE, JFIF, TIF, TIFF                                                     
                                                                                                                 

Specifications subject to change without notice.

Specifications

Module NameSLM VISAI-S100
Operating SystemWindows 10 (64 bit)
CPUMinimum: Intel Core i5 / Recommended: Intel Core i7
GPUMinimum: Nvidia GTX 1070 (RAM: 8GB) / Recommended: Nvidia RTX2080 Ti
RAMMinimum: 16G / Recommended: 32G
InterfaceCamera Interface
Coding InterfaceNet framework 4.8
Coding LanguageC #
Compatible WithMVTec Halcon, NI LabVIEW
LanguageEnglish
Image FormatPNG, BMP, JPG, JPEG, JPE, JFIF, TIF, TIFF                                                     
                                                                                                                 

Specifications subject to change without notice.

Solvision Learn about Solvision products

Solvision

Learn about Solvision products

Solvision Learn about Solvision products

Solvision

Learn about Solvision products

DO YOU HAVE ANY QUESTIONS?

SOLOMON personnel is always ready to assist you.

DO YOU HAVE ANY QUESTIONS?

SOLOMON personnel is always ready to assist you.