AT logo - The Machine Vision Specialists  
  Cameras  
  Frame Grabbers  
  Lenses  
  Lighting  
  Software  
  Vision Systems  
  Accessories  
  News  
  Careers  
  Contact AT  
  Site Map  
Request product information

Contact
info@adeptturnkey.com.au

Perth:
(08) 9242 5411

Sydney:
(02) 9905 5551

Melbourne:
(03) 9384 1775


Defence Recognised Supplier Scheme Logo

 

Sapera Software Suite: an intuitive, easy-to-use software package for Machine Vision

Back to Newsletter
Go to Software

Sapera Vision Software Edition 2022-05 for machine vision applications

Sapera Vision Software provides image acquisition, control, image processing, image analysis and artificial intelligence functions to design, develop and deploy high-performance machine vision applications. Sapera Software is designed for a broad range of applications, from precision metrology to pharmaceutical packaging. With an easy to use interface, Sapera provides a suite of tools that can be readily applied to a multitude of automated inspection tasks, such as positioning, measuring, identification and flaw detection.

 




 
What is new in Edition 2022-05 of Sapera Vision Software?
Astrocyte 1.30

Astrocyte is a code-free AI training tool to quickly deploy AI models for machine vision solutions. It allows users to use their own images of products, samples, and defects to train neural networks to perform a variety of tasks such as anomaly detection, classification, object detection, segmentation, and noise reduction. With its highly flexible graphical user interface (GUI), Astrocyte allows visualizing and interpreting models for performance and accuracy as well as exporting these models to files that are ready for runtime in Teledyne DALSA Sapera and Sherlock vision software platforms.

In this new version of Astrocyte (1.30), the following key features have been introduced:
Tiling on Large Images
Tiling is a mechanism for handling large images with small details. Tiling allows applying training and inference on native resolution without using intensive memory and without downscaling and causing loss of details. The tiling workflow consists of decomposing the input image into sub-images (tiles), iteratively processing the tiles at native resolution, and recombining the tiles at the output. The tile size is selected either automatically or manually.

Below is an example where tiling is used for locating knots on wood planks. The input image has a resolution of 2800x1024 and contains knots with sizes as low as 10 pixels. Tiling allows preserving the native resolution to locate those small knots with precision (instead of scaling down the input image and causing the knots to disappear).

Increased Performance on Anomaly Detection
Astrocyte introduces a new Anomaly Detection algorithm providing better performance on low to medium resolutions compared to the previous version. Use Anomaly Detection directly on images ranging from 256x256 to 1024x1024. For resolutions above 1024x1024 you can combine Anomaly Detection with tiling or decimation. This new algorithm replaces the old “pixel-level” algorithm and is the new default. The old algorithm is preserved for legacy purposes.

Anomaly Detection does not need annotations at training, making it easier to train than object detection and segmentation. Anomaly Detection can locate any defect never seen at training given the training is performed on good samples only. Below is an example of Anomaly Detection usage. On the left, is a normal image with no defects. In the middle and on the right, are different types of defect shapes (highlighted with pseudo-color). The challenge in this example is to differentiate the defects from the normal lettering.

YOLOX Object Detection
Astrocyte now offers an additional object detection algorithm called YOLOX. It is more compact and in general provides better performance than the existing SSD (Single Shot Detector) algorithm. Because of its compactness, YOLOX is widely used in embedded devices. SSD is kept for legacy purpose. In the example provided with Sapera Processing (“hardware dataset”), YOLOX increases the accuracy from 95% to 99% while running 50% faster compared to the previous SSD algorithm.

Processing of Non-Square Images
Astrocyte now uses a flexible mechanism for preparing the input image before training and inference. The image sent to the neural network can have different width and height. This ensures preserving the original aspect ratio (and, therefore, avoid distorting the image with stretching or padding). At training time, you have the option of using either the original, decimated, or tiled input image. Below are examples of the new (left) and old methods (middle and right).

Other General Improvements
Now supports frame-grabbers allowing Astrocyte to acquire live images from any Teledyne DALSA device
New diagnostic report to help troubleshooting training issues at customer sites
Dynamic discovery of cameras (automatic detection of plug/unplug GigE/USB3 cameras)
Saving datasets as TAR format to allow altering and merging previously created datasets
Improvement of training progress display (now displays both epoch numbers and steps)
Improvement of brush tool for segmentation annotations

Click here to find out more about Astrocyte:

Sapera Processing 9.30.

Sapera Processing is the cornerstone of the overall Sapera Vision Software platform. Sapera Processing is an extensive library of image processing, image analysis and AI functions dedicated to application development, integration, and deployment. In version 9.30 of Sapera Processing new functions and improvements are introduced mostly on the AI and 3D tools.

Click here to find out more about Sapera Processing:

 

Need a price or more application information? Please email Adept Turnkey or call our offices
Adept Turnkey Pty Ltd is"The Machine Vision and Imaging Specialists" and distributor of Teledyne Dalsa products in Australia and New Zealand.
To find out more about any Teledyne Dalsa product, please call Adept Turnkey at Perth (08) 9242 5411 / Sydney (02) 9905 5551 / Melbourne (03) 9384 1775 or contact us online.

 

 

If you like this page, please recommend and share it.

Google+ Facebook Twitter More