Video tone mapping for a XIMEA 8.9 MPix camera on the GPU
Updated:
Author: Fyodor Serzhenko
Tags: XIMEA, machine vision, Tone Mapping, ALTMapper, CUDA, Camera ISP on GPU, video
This page shows real-time tone mapping applied to footage from a XIMEA 8.9 MPix industrial camera (Sony IMX255, USB3) processed by a GPU-based ISP. ALTMapper local tone mapping recovers detail in highlights and shadows and stays temporally stable across frames — no flicker. The clip below shows the result directly, so you can judge the quality yourself.
Video from a XIMEA 8.9 MPix camera with a GPU-based ISP: tone mapping applied on the left, the same pipeline without the tonemapper (sRGB gamma only) on the right.
Detailed benchmark data. Measured results are published on our benchmark pages: Tone-mapping benchmarks, all benchmarks.
See it on your own images
Send us a few of your own frames — RAW, TIFF, JPEG or PGM — and we'll run ALTMapper on them, so you can judge the result on your own data.
What you are seeing
The camera captures a wide dynamic range scene in RAW. Both halves run through the same GPU ISP (demosaicing, colour, gamma); the only difference is the tonemapper. On the left, ALTMapper adaptive local tone mapping is applied — shadows and highlights keep their detail and local contrast. On the right, the same pipeline runs with sRGB gamma only, so bright and dark regions collapse. Watch a moving part of the frame: the tone-mapped side stays stable, without the brightness pulsing or shadow “crawl” that per-frame tone mapping produces.
Why it matters for XIMEA machine-vision cameras
Industrial cameras like the XIMEA 8.9 MPix (Sony IMX255) capture far more dynamic range than a display can show. In a live machine-vision or inspection pipeline the RAW-to-RGB conversion has to run in real time, and tone mapping has to fit inside that budget without introducing flicker or motion artifacts. Running ALTMapper on the NVIDIA GPU keeps the conversion real-time and deterministic, so what the operator sees is stable frame to frame.
How it works
ALTMapper is Fastvideo’s adaptive local tone mapping engine on NVIDIA GPU: a per-pixel transform built from global and local image features, deterministic and AI-free, fast enough to run inside a real-time camera ISP. Why per-frame local tone mapping flickers on video — and how temporal stability is achieved — is explained, together with the algorithm and single-shot HDR, on the local tone mapping on GPU page.
See it for yourself
The demo above is the point: judge the quality with your own eyes. You can also run ALTMapper on your own footage — download the ALTMapper software (free demo). Performance numbers for the current build are being finalised and will be published on the product page.
Related
- Local tone mapping for ARRI LogC cinema footage
- ALTMapper: adaptive local tone mapping on GPU
- FastVCR software on GPU for XIMEA cameras (Windows/Linux/ARM)
About the author
Fyodor Serzhenko, PhD, is the founder and CEO of Fastvideo. He earned his PhD at the Moscow Institute of Physics and Technology (MIPT) in 1993. Since 2009 he has led the development of Fastvideo’s GPU-accelerated image codecs and ISP modules, including the Fastvideo SDK. Connect on LinkedIn.
Why you can trust these results
Fastvideo has built GPU-accelerated image processing software since 2009. Our codecs comply with the standards they implement — JPEG (ITU-T T.81 / ISO IEC 10918) and JPEG2000 (ITU-T T.800 / ISO IEC 15444) — and our tools are open source on GitHub. Every performance figure we publish is reproducible: see our benchmark methodology, download the Fastvideo SDK benchmark report (PDF), and measure it on your own GPU.