
Image Pro AI
AI-powered image segmentation with deep learning for accurate object detection in life sciences and materials research.
Overview
Image Pro AI is a 64-bit scientific image analysis platform developed by Media Cybernetics. It incorporates deep learning technology to perform image segmentation, and is designed for researchers and scientists working in life sciences as well as materials research and inspection. The platform aims to reduce the time spent configuring analysis pipelines by providing intuitive Analysis Protocols and pre-trained AI models.
Image Pro AI supports deep learning model architectures including Cellpose, StarDist, and UNET, and includes a library of pre-trained models covering a wide range of 2D image analysis applications. All processing runs on-device, meaning no data is uploaded to or downloaded from external servers.
Key Features
- Superior outline accuracy: AI models locate and trace objects regardless of dim or complex edges, handling any shape of object or phase with intricate detail, including touching or overlapping objects.
- User-friendly operation: Deep learning segmentation requires minimal user input, with simple controls, limited options, and fast prediction results.
- On-device processing: All operations run locally, with no uploading or downloading required, keeping data secure and confidential.
- Protocol-ready automation: Deep learning segmentation can be integrated into automated workflows by saving an Analysis Protocol with either a custom or pre-trained model.
- Multi-channel support: Powered by AI machine learning with multi-channel designs that are customizable across similar routines for repeatable performance and parallel fast analysis.
Deep Learning Modules and New Features
- 2D AI Deep Learning Module
- AI Deep Learning Prediction for segmentation
- AI Deep Learning Trainer for training custom models
- AI Deep Learning Model Browser
- Support for Cellpose, StarDist, and UNET model architectures
- Most protocols updated to support AI deep learning
- EM Protocols based on AI deep learning
- Expanding library of pre-trained AI deep learning models
Pre-Trained Model Library
The ready-to-use model library covers a broad range of 2D image analysis applications, including:
- Fluorescent Cells
- Fluorescent Dendrites
- Fluorescent Nuclei
- Fluorescent Overlapping Nuclei
- Grains: Titanium (Ti6Al4V)
- Grains: Titanium (Ti6Al4V) Alpha Phase
- Hematoxylin Nuclei
- Label Free Cells
- Retinal Blood Vessels
- TEM Mitochondria
- TEM Nuclei
- Versatile Grains
- Versatile Particles
- Widmanstatten Patterns
Image Pro AI is categorized under Media Cybernetics products and is positioned as the successor to earlier Image-Pro platforms. Its on-device architecture addresses data confidentiality concerns by avoiding reliance on external or shared company servers, making it suitable for environments where data security is a priority.
