Display of EVS software application interface and supporting hardware (camera and edge box) designed for embedded AI series of products developed by Aqrose Technology

Industrial AI Vision Algorithm Platform

Aqrose AIDI is the platform software for industrial applications. It is designed with 8 AI functions for various applications, including defect positioning, detection, classification, character recognition, etc. Our graphical interface allows an all-in-one AI model with image annotation, model training, and tuning without coding. Our software supports integration into diverse production line devices from various industries, meets distinct detection needs in different applications, and benefits users with quality control and yield rate improvement.

Product Strength

Strong performance and wide coverage of scenarios
Diverse functionality and high deployment efficiency
Flexible Secondary Development and low cost

Solve various complex detection problems

Effectively deal with complex detection scenes such as low contrast, various types of defects, complex background, and small defect scale.The minimum detectable size is 3 pixels, and can be as low as 2 pixels for defects with significant contrast.

Adapt to high-speed production lines

It takes about 10min to train 100 images, less than 30ms for single image pixel-level defect detection, and less than 1ms for defect classification.

Support small sample scenarios

Built-in small sample algorithm, which supports unsupervised OK image learning mode, and accurately identifies single-class defects with only 30-50 images, or even without defect samples.

Features

SEGMENT

Pixel-level detection on images to precisely recognize defects’ location, size and category.

ANOMALY SEGMENT

Performs pixel-level detection on all known and unknown defects with a reference image of the defect-free product for rapid deployment and verification.

DETECT

Designed to perform region-level detection on the images to accurately identify objects or defects.

LOCATE

Detects category, position, size, and orientation of single or multiple objects within an image.

OCR

Designed for accurate location and efficient recognition of characters in images.

ASSEMBLY VERIFY

Designed to detect if the presence and quantity of objects in a specific region meets the requirement.

CLASSIFY

Designed to determine the category of the entire image.

ANOMALY CLASSIFY

Performs whole-image classification on all known and unknown defects with just a reference image of the defect-free product for rapid deployment and verification.