Abstract: Counting, measuring, and identifying particles is a crucial aspect of various research endeavors. Typically, images containing particles are manually processed using a software ruler. Automated processing techniques, which rely on conventional image processing methods such as edge detection and segmentation, are not universally applicable and require setting several parameters through trial and error. Additionally, these techniques can only be utilized on high-quality images. Also, the ambiguity of the data set can greatly affect the quality of object identification. The report presents the iOk platform (iok.nsu.ru), which uses artificial intelligence through the ParticlesNN web service and Telegram bots DLgram and No Code ML as well as other means of detecting objects on an image. The platform provides automatic search and analysis of objects in images without pre-processing, regardless of the type and quality of the image. At the output, you can obtain information about object recognition, its area and size, as well as its position in the image. The neural network can be trained on user images, no programming skills are required. All the services are free.

Cite: Kudinov V.Y. , Matveev A.V. , Nartova A.V. , Sankova N.N. , Belotserkovskii V.A. , Okunev A.G.
iOk Platform for Automated Object Detection and Analysis in Microscopy Images
In compilation 2023 IEEE XVI International Scientific and Technical Conference Actual Problems of Electronic Instrument Engineering (APEIE). 2023. – C.1420 - 1423. DOI: 10.1109/apeie59731.2023.10347794 Scopus OpenAlex

Identifiers:

Scopus:	2-s2.0-85182274888
OpenAlex:	W4389667029

Citing: Пока нет цитирований

Altmetrics: