ALTMapper: Adaptive Local Tone Mapping Software
Updated:
ALTMapper is adaptive local tone mapping software that processes a 65 MPix (9344×7000) image in 6.3 ms — about 10 GPix/s — on an NVIDIA GeForce RTX 4090, and runs the same deterministic algorithm on NVIDIA Jetson and on CPU. It recovers shadow detail, controls local contrast, and removes haze with 32-bit ISP accuracy using a classical per-pixel transform — no neural networks and no hallucinated detail, so the result is deterministic and repeatable. ALTMapper ships as a GUI application for interactive work, a CLI application for batch and automated processing, and a CUDA API for embedding into GPU pipelines; the underlying method is described on the local tone mapping page.
Detailed benchmark data. Measured results are published on our benchmark pages: Tone-mapping benchmarks, all benchmarks.
ALTMapper Features
- Visual perception image enhancement with controlled contrast preservation: global, local, and micro contrast
- Image detail preservation
- Shadow recovery and local gain control; Over Gain mode for low-illumination cases
- Saturation correction for better visualization
- Moderate fall-off compensation
- Soft cloud shadows suppression
- Dehazing
- Automatic histogram smoothing
- Flicker reduction for video tone mapping
- 360 panoramas support
- Consistent quality across rural and urban environments
- No hallucinations (no AI): deterministic, classical algorithm
- ISP with 32-bit accuracy on CPU and GPU
- Built-in timing and performance measurements
See the difference
Source image on the left, ALTMapper result on the right — recovered shadow detail and controlled local contrast, with no halos and no hallucinated detail.
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.
ALTMapper Benchmarks
How fast is adaptive local tone mapping on the GPU?
GPU tests: 16-bit per channel RGB image, all data in GPU memory, timing includes GPU computations only.
| Platform | Image | Time | Throughput |
|---|---|---|---|
| NVIDIA GeForce RTX 4090 (CUDA 12.6) | 9344×7000 (65 MPix) | 6.3 ms | 10 GPix/s |
| NVIDIA Jetson AGX Orin 32GB (JetPack 6.2.1) | 4128×3008 (12 MPix) | 5.4 ms | 2.3 GPix/s |
These numbers matter for workloads with large volumes of images or continuous video: the CUDA implementation was designed for exactly such heavy-load tasks, while the CPU version covers systems without a GPU.
Compatibility
- OS: Windows 10/11, Linux Ubuntu, Linux4Tegra (L4T)
- NVIDIA GPUs (GeForce RTX, NVIDIA RTX) with CUDA 12.6
- NVIDIA Jetson: Nano, Xavier NX/AGX, Orin NX/AGX, Thor
- Input formats: HDR and high bit-depth sources, JPEG, TIFF, BMP, PNG, PPM/PGM
- Delivery: GUI application, CLI application with Python control, API for GPU pipelines; also available as a module of the Fastvideo SDK
Download
Applications
- Machine vision and industrial imaging
- Street view, mobile mapping, and aerial imaging
- Digital photography and digital cinema
- Medical imaging
- Broadcast and streaming
- Display devices and computer vision
ALTMapper Roadmap
ALTMapper is a mature tone-mapping engine in active development. Delivered in recent releases (2025–2026):
- HDR engine on GPU for raw, color, and monochrome images
- Local tone mapping for video HDR
- Local tone mapping library on the CPU
- Performance optimization; image quality improvements
Current development focuses on:
- Local tone mapping for machine vision cameras with dual gain ADC
- Support for high dynamic range image sensors
ALTMapper FAQ
What is adaptive local tone mapping?
It is a per-pixel, spatially adaptive transform that compresses a high dynamic range into a displayable range while preserving local contrast and fine detail. Unlike a single global curve, ALTMapper adapts to each region of the image, so it lifts shadows and tames highlights without flattening texture.
Does ALTMapper use AI or neural networks?
No. ALTMapper is a deterministic classical algorithm and does not invent or hallucinate detail. The same input always produces the same output — important for machine vision, measurement, and broadcast pipelines that need repeatable results.
How fast is ALTMapper on the GPU?
On an NVIDIA GeForce RTX 4090 it processes a 65 MPix (9344×7000) image in 6.3 ms — about 10 GPix/s — timing GPU computation on a 16-bit RGB image held in GPU memory.
Does it run on NVIDIA Jetson?
Yes. On a Jetson AGX Orin 32 GB it processes a 12 MPix image in 5.4 ms (about 2.3 GPix/s), and it supports the Jetson line — Nano, Xavier NX/AGX, Orin NX/AGX, and Thor — under Linux4Tegra.
Can ALTMapper process video?
Yes. It includes flicker reduction for video tone mapping, so brightness and contrast stay temporally stable from frame to frame.
Is there a CPU version?
Yes. The library runs on CPU with the same 32-bit ISP accuracy for systems without a GPU; the CPU path covers full I/O including SSD read/write and JPEG decode and encode.
How can I try it?
Download the free ALTMapper demo for Windows and measure the speed on your own hardware. ALTMapper is also available as a module of the Fastvideo SDK.
Related: how the underlying local tone mapping algorithm works on GPU, and side-by-side video demos on ARRI Alexa LogC cinema footage and a XIMEA 8.9 MPix camera.
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.