Deep learning vision software automates visual inspection applications
SuaKIT vision software performs visual inspection and classification of complex materials and key components for the mobile, semiconductor, electronics, and automotive industries. It also automates manual inspections in the food and beverage, packaging, and raw materials industries.
Improves quality and yield
SuaKIT’s finetuned deep learning models deliver highly accurate inspection results. The deep learning algorithm’s internal analysis process enhances quality upstream to reduce overkill and underkill rates to optimize quality and yield.
Implementing an automated system decreases dependence on unreliable, manual inspections. Being able to run inspection operations around the clock optimizes throughput and improves takt time to satisfy customer demand. SuaKIT’s high detection rates also reduces the need to addition and costly inspection hardware.
Ensures reliable, validated results
SuaKIT’s highly consistent inspections guarantee the same results from line to line, shift to shift, and factory to factory. The software archives images and document results that can be reviewed and validated offline. This valuable data helps quality engineers to optimize applications and understand anomalous results.
Detects objects within different classes in a single image
Groups images by multiple predetermined classes
Accurately finds position/area/shape or defects in the images
Deep learning architectures
Single Image Analysis
Learns each image and identified defects
Learns and detects defects by concentrating on the differences between a set of two images
Multi Image Analysis
Analyzes relationships between images to train defect detection model
One Class Learning
Identifies defects based on deviations from trained OK images