MediaLab
MediaLab
Rating (5)
Reviews: 1
Category: Multimedia design
Package 17 has pre-built chromakey as well as new ROI and CV bitwise and arithmetic operations for all your matting needs!  Also showing off the new Grid layout in Pipeline Outputs panel.
Package 16 has the new Live Update mode, provides interactive feedback as you adjust parameters via the sliders.
Package 15 Histogram Data operation provides access to normalization as well.
Package 15 new Histogram operations: Draw Histogram, Histogram Data, and Histogram Equalization.
Package 15 there is the new Contrast Limited Adaptive Histogram Equalization operation.
Package 10 the Pipelines page provides a better playground.
Run ONNX models via Windows ML to generate images.  Here we are doing "style transfer" with mosiac.
Use Azure Cognitive REST APIs for vision tasks, like "analyze" or "ocr".  Saves all responses in original JSON with metadata.
Combine a folder of images into a new video.
Directly process an entire video with an OpenCV operation.

Description

We built this app to help process videos from field research experiment, and now we have made it available to everyone. They were using other open source video processing software, and it was failing for them.

The main feature of MediaLab is access to OpenCV, Windows ML (ONNX), and REST APIs for processing video and images, without having to program in Python or C++ by hand.

For our task, it was contour extraction, and needed to generate CSV contour data from raw (infrared) video footage, which were literally Days Long!

If you are involved in processing videos or images from research, or just interested in understanding Computer Vision techniques without any programming, then have a look at MediaLab!

Recent updates include other image processing operations like chromakey matting, bitwise operations with masking, and region-of-interest (ROI) extract and combine.j

Live Update mode shows you the entire pipeline while you adjust operation parameters.

  • Extract video frames to a folder
  • Analyze folders of images with OpenCV operations
  • Save OpenCV configurations as JSON "pipelines" for later use
  • Batch process folders of images with a pipeline
  • Batch process an entire video with a pipeline
  • Combine folder of images back into a video (i.e. slideshow)
  • Multi-step pipelines can save CSV data and images (depending on operation)
  • Interactively explore the Operation Library on images with Preview
  • Use ONNX models for classification (in CSV) and image creation
  • Use Azure Cognitive REST APIs to capture JSON responses for offline analysis
  • Over 40 operations spanning OpenCV, WindowsML and REST
  • Perform chromakey and other matting tasks
Product ID: 9PP8Q6XVH1BQ
Release date: 2020-11-02
Last update: 2022-12-22