RAW Capture
RAW10/12, Bayer pattern, black level, dark current, defective pixel maps.
ISP Solutions
ISP Solutions
Starting from real samples, target platforms, and image quality issues, Tansontech covers RAW/YUV, HDR/LTM, 3A, BNR/TNR, Color, Infrared, Embedded, and related ISP/IQ algorithm areas with integrable imaging algorithms, reference implementations, and tuning support.
Pipeline Capability Scope
The ISP solution covers key processing nodes from RAW input to RGB/YUV output and provides algorithm integration support around target platforms, image formats, and calling interfaces.
Preprocess
BLC, LSC, BNR, Demosaic, noise behavior, and texture protection.
AE / AWB / AF
Metering, white balance, focus metrics, temporal stability, and control state machines.
Merge + PWL
DOL, DCG/DAG, Sensor HDR, inverse PWL, and ghost suppression.
Local Mapping
Dynamic range compression, halo control, local contrast, and natural rendering.
Color + YUV
CCM, Gamma, sharpen, YUV420, and display matching.
Solution Modules
These modules are common capability examples. Each can be validated separately or integrated into a full pipeline, then extended around sensor, lens, target platform, scene, and image quality goals.
HDR / LTM
Dynamic range and local mapping
Multi-exposure merge, inverse PWL, local contrast, halo control, and natural rendering.
BNR / TNR
Noise and detail
Bayer-domain noise models, spatial-temporal denoise, texture protection, sharpening, and artifact control.
3A
Auto control
AE, AWB, AF, flicker handling, mixed illuminants, and temporal stability strategies.
COLOR
Color science
CCM, Gamma, Tone Curve, saturation, batch consistency, and style matching.
INFRARED
Infrared imaging
NUC, defective pixels, DDE, pseudo color mapping, detail enhancement, and preprocessing.
INTEGRATION
Algorithm integration
Reference implementation, interface notes, parameter configuration, test samples, and platform integration support.
Engagement Process
ISP algorithm cooperation starts by clarifying data, platform, and goals, then moves into prototype validation, algorithm implementation, and integration support.
Samples and requirements
Confirm product type, sensor, lens, chip platform, RAW/YUV/RGB data, and target scenes.
Issue diagnosis
Analyze exposure, dynamic range, noise, color, sharpening, local contrast, and temporal stability.
Algorithm prototype
Use Python/C++ to validate algorithm direction, parameter range, effect ceiling, and key risks.
Integration support
Provide algorithm implementation, interface notes, parameter configuration, test samples, and integration support.