3. ISP System Overview

3.1. Function Overview

The ISP system supports standard image processing functions, including basic functions such as defective-pixel correction, lens shading correction, auto exposure, auto white balance, auto focus, and demosaicing. It also supports advanced processing functions such as noise reduction, WDR, and DRC.

The main image processing functions supported by the ISP are as follows:

  • Cropping (Crop)

  • Black Levelcorrection (BLC)

  • wide dynamic rangeFunction (WDR)

  • Static and dynamic defective-pixel correction (DPC)

  • Crosstalk removal (CrossTalk Removal)

  • Bayer noise reduction (BNR)

  • Lens shading correction (LSC)

  • Dynamic range compression (DRC)

  • Demosaicprocess

  • Purple-fringe correction (PFR)

  • Color correction (CCM)

  • Gammacorrection

  • Automatic dehazing (Dehaze)

  • 3D color lookup-table enhancement (CLUT)

  • Local contrast enhancement (LDCI)

  • Global contrast enhancement (DCI)

  • Image sharpening (Sharpen)

  • 3D noise reduction (3DNR)

  • Chroma noise processing (CNR)

  • Auto exposure (AE)

  • Auto focus (AF)

  • Auto white balance (AWB)

  • 3A-related statistics output

  • Brightness shading support

3.2. Functional Block Diagram

The overall ISP structure is shown in Figure 3.1 as shown.the relevant informationdocumentationofnextsectioncontentwilldescribethe relevant informationmoduleofFunction, module (BLC, LSC, AWB, CCMandCLUT) ofParametercalibrationmethodandimagequalityTuning Method.

_images/ISPSys002.png

Figure 3.1 Overall ISP structure

3.3. Module Overview

the relevant informationdocumentationofnextsectioncontentwilldescribethe relevant informationmoduleofFunctionandimagequalityTuning Method.ISP module functionsthe relevant informationas Table 3.1 as shown.

Table 3.1 ISP module functions

Module Name

Function

Crop

Crops the input image.

BLC

Black Levelcorrection.

WDR

the relevant informationfusionofwide dynamic rangeFunction.

DPC

Detects and corrects static and dynamic defective pixels.

CrossTalkRemoval

Corrects the imbalance between the Gr and Gb channels.

BNR

Performs image denoising in the Bayer domain.

LSC

Provides lens shading correction.

DRC

adjustimageofdynamic range, the relevant informationindisplaydeviceaboveofdisplayeffectandthe relevant information.

Demosaic

Converts Bayer-format raw images to RGB images.

PFR

Removes purple fringes and improves purple-fringe artifacts at image edges.

CCM

Uses a 3x3 matrix to correct colors.

Gamma

Adjusts overall image brightness according to the gamma curve.

Dehaze

Performs dehazing for hazy scenes and improves image contrast and sharpness.

CLUT

Uses a 3D LUT for complex color adjustments, including brightness and saturation adjustment.

CSC

Converts RGB images to YUV images using a 3x3 matrix and vector offsets.

LDCI

the relevant informationinforimagethe relevant informationstatisticsofmethod, enhancementimageofthe relevant informationcontrast, samewhencanadjustmentfilteringParameter, the relevant informationadjustthe relevant informationcontrastenhancementofthe relevant informationrange.

DCI

the relevant informationinhistogramthe relevant informationofmethodthe relevant informationhighoverallimageofcontrastanddark areaofdetail.

Sharpen

the relevant informationimageofsharpeningFunction, increaseimagesharpness.

3DNR

Removes image noise through temporal filtering, preserves image detail, and reduces the encoding bitrate.

CNR

Removes chroma noise from images and reduces color blotches.

CA

Provides saturation adjustment and thermal-imaging colorization.

CA_Lite

Provides saturation adjustment.

AE

Provides auto-exposure statistics to software for Sensor adjustment and automatic exposure.

AWB

Provides global and regional statistics to software for Sensor adjustment and automatic white balance.

AF

This module outputs image-sharpness statistics, which software uses to perform auto focus.