INTRODUCING THE EyePASS
The Eyecon2 our flagship technology, used to measure the particle size and shape of powders and bulk solids. The Eyecon2 is a direct imaging, non product contact, particle analyser that reports data of your bulk powder processes in real time.
More info here for those of you that are new!
One of the primary steps performed by our image analysis software EyePASS is particle segmentation i.e., the detection and delineation of particles from the images sent by the Eyecon2. All measurements reported by the software are based on the results of the particle segmentation step. To date EyePASS 2.0 uses traditional image-processing techniques to achieve the segmentation.
The EyePASS Artificial intelligence (AI) project is using cutting edge machine learning (ML) techniques to drastically improve the ability to segment particles, enhanced segmentation means a wider range or material types can be detected with superior accuracy in measurement results.
The machine learning approach utilises convolutional neural networks (CNN), It is a supervised machine learning method where learned features with input data are manually labelled for the task of identifying particles. To break that down a little more, EyePASS AI images of materials are labeled manually by in-house experts, then a CNN iterates and improve its feature extraction of the images. The features are not trained but are instead learned by the network. More and more datasets are been prepared as part of the project to challenge and improve the model.
The screenshot shows some recent results from the ongoing work.