Projects in Development

Aside from developing turnkey products and customized solutions we also carry out research in the field of image and video processing, including fast JPEG/JPEG2000/Bayer compression and decompression, image demosaicing, denoising, filtering, HDR, video processing and much more. You can find the list of completed projects below. Please don’t hesitate to contact us regarding any of them.

Open source projects from Fastvideo

• GPU Camera Sample application - https://github.com/fastvideo/gpu-camera-sample for realtime GPU-based image processing for machine vision cameras from XIMEA, Basler, Baumer, EVT, JAI, FLIR/DALSA, Imperx, Matrix-Vision, Lucid Vision Labs, IO Industries, Daheng Imaging, Mindvision on Windows/Linux/ARM
• PGM2DNG converter - https://github.com/fastvideo/pgm2dng

Current projects

• FastVCR software for EVT cameras - soon
• J2K RAW encoding with minimum loss and compression ratio in the range of 3-6 times - soon
• FastVCR CLI application with remote control - done
• FastVCR software for XIMEA cameras (realtime raw image processing on GPU with integrated camera control) - done
• Fast J2K Player - done
• Highlight recovery on GPU - done
• NVIDIA Jetson Orin support - done
• Support for IOI cameras: Volucam, Redwood, Flare, Victorem - done
• Fast J2K Viewer - done
• Fast image stitching on GPU with BigTIFF and OME-TIFF support for microscopy applications - done
• Accelerated FFC module - done
• Local and Global Tone Mapping on GPU for HDR images - done
• High performance RAW to JPG conversion on GPU for Teledyne DALSA Genie Nano 5GigE C8100 color camera (On-Semi XGS45000 image sensor) - done
• Image processing pipeline (RAW to RGB) on Jetson for Leopard Imaging mipi csi camera with 12-bit SONY IMX477 image sensor - done
• Dual ADC (combined and non-combined) image processing for SONY Pregius Gen4 image sensors - done
• Image processing pipeline (RAW to RGB) on Jetson for 24-bit automotive OnSemi AR0820 image sensor - done
• HDR processing for automotive SONY IMX490 and IMX728 image sensors - done
• Chromatic Aberration Removal module in RAW and/or RGB domains - done
• High performance JPEG XS decoder on GPU - done
• High quality undistortion calibration without radial model - done
• CUDA-12.1 support - done
• Disparity map computation on GPU for stereo cameras - in progress
• Accelerated compact Remap module - in progress
• Image processing software for retinal surgery - in progress
• Camera stitching on GPU
• Video stabilization on GPU
• GPU memory manager

Software which is based on Fastvideo SDK

• FastVCR software for realtime capture, processing, encoding and streaming
• MXF Player
• J2K Viewer
• Remote color grading software
• Fast CinemaDNG Processor (GUI and server applications)

FFmpeg filters and codecs from Fastvideo

• Video transcoder with internal Remap on GPU
• JPEG2000 decoder for FFmpeg
• JPEG2000 (MXF) live transcoding to H.264 or H.265
• Fast format transforms on CUDA for FFmpeg applications

Completed projects

• SDK for fast JPEG library on CUDA
• SDK for fast GPU debayer library
• Combined optimized solution Debayer + JPEG
• 16-bit debayering on CUDA (L7, DFPD, MG)
• 16-bit image and video processing library on CUDA
• 12-bit JPEG codec on CUDA
• Support for HD-SDI grabbers
• Realtime image processing on Jetson TK1, Jetson TX1 and Jetson TX2, Jetson AGX Xavier
• Fast discrete wavelet transform on CUDA
• Lossy and lossless JPEG2000 image encoding and decoding on CUDA
• Bayer Codec
• 3D LUT Transform on GPU
• Color management with ICC profiles on GPU
• Full image processing pipeline on GPU for high speed and high resolution cameras
• Fast and high quality image and video resizing
• Bad pixel removal on GPU
• Fast Gauss Blur for image processing (High Pass and Low Pass filters)
• High quality histogram equalization and contrast enhancement
• Camera RAW image and video processing on GPU
• Image denoiser on CUDA
• Median filter on GPU
• Fast MJPEG codec on GPU
• Image trigger for high speed cameras
• Long term high speed video recording system (hardware, firmware and software)
• DICOM image processing on GPU
• Fast Lossless JPEG Decoder on CPU