Push your camera to the Dotphoton level
Validated by imaging specialists
The Dotphoton customer validated image compression and optimization algorithm will allow you to improve the performances of your camera, to make the end-user's life easier by keeping the same workflow and to deliver images compatible with all existing image processing algorithms and new Artificial Intelligence and Machine Learning algorithms.
How Dotphoton works:
Typically, the quality metric for image compression has been the difference between the original image and the compressed image. The original images however contains a high degree of noise, which is hard to compress, resulting in a low compression ratio, or loss in quality.

Dotphoton's figure of merit for quality, on the other hand, is the difference between what a perfect sensor would have measured, and the compressed value.

This gives Dotphoton a way in rearranging the noise so that it can easily be compressed, allowing it to outperform traditional compression methods both in terms of quality and compression ratio.
Benefits
Transfer 10 times for images with your SUB - 1Gbit E
Data storage cost divided by 10
Enable Ai on RAW
Same workflow thanks to existing file format
Same workflow thanks to existing file format
Long term preservation
Applications
Science and research
Industrial inspection
Medical and healthcare
Automobile
Mobile
Dotphoton compression parameters
Image
Performance
Quality
Input formats
Your RAW format open or proprietary
Output formats
● .TIFF
● .PNG
● .DNG
●JPEG2000
●.dcmPixel format: raw greyscale, raw CFA (RGB, quad, x-trans etc)
● multispectral and hyperspectral (in development)
Bits-per-sample
Up to 16 bits (≥12 bits recommended)
5:1 – 13:1

Typical compression ratio
2:1

Min compression ratio
~1Gbps / core

~1Gbps / core
1 line readout (~ 5μs)

Hardware latency
100 ops/pixel

Complexity
Memory

base: 1-line + 1kB
all features: 3-frames + 128kB.
Information loss
< 0.3 bits per pixel
Difference w.r.t. original sample
< 1σ
Artefacts
No block artefacts
No posterization
No temporal artefacts
No ringing
No contouring
No aliasing
No low-pass filtering.
Bias, systematic errors
< 0.1 photons
Requirements
1
Gain > 1DN/e-
2
Negligible loss of information prior to
compression
3
Camera hardware available for
characterization


Optional features
Pixel correction
Data integrity verification
Linearization (global)
Camera authentication
Linearization (per pixel)
Uncertainty metadata
Compatibility: tested on more than 50 cameras from a dozen manufacturers
1
sCMOS, CCD sensors
2
Rolling shutter
3
Global shutter
-4-
.tiff or any lossless open format
Licensing for:
Dynamically library (dll)
for Windows, Linux and Mac, and has been tested with C, C++, Objective-C, Python, C# and Swift (others may work)
FPGA IP
Customizable
Virtual machine
Depoyable on vmware to dynamically compress images on network shares
Graphical user interface
For Windows, Linux and Mac
Command-line utility
For Windows, Linux and Mac

Be the first to test our next products and learn about new features