3. ISP System Overview

3.1. Function Introduction

ISP system supports standard image processing functions, including bad point correction, lens shadow correction, auto exposure, auto white balance, auto focus, demosaic and other basic functions, as well as advanced processing functions such as noise reduction, WDR and DRC.

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

• Black level correction (BLC)

• Static and Dynamic Defect Pixel Correction (DPC)

• CrossTalk Removal

• Lens shadow correction (LSC)

• Bayer noise reduction

• Demosaic processing

• Color Correction Matrix (CCM)

• Gamma correction

• Purple fringing correction (LCAC, RGBCAC and CAC)

• Wide Dynamic Range (WDR)

• Dynamic Range Compression (DRC)

• Auto Exposure (AE)

• Auto Focus (AF)

• Auto White Balance (AWB)

• 3A related statistical information output

• Image Sharpening (Sharpen)

• Auto dehaze treatment (Dehaze)

• Local Dynamic Contrast Improvement (LDCI)

• 3D noise reduction (3DNR)

• Color 3D look up table enhancement (CLUT)

• Brightness coloring

• Digital image stabilization

3.2. Function Block Diagram

The overall structure of ISP is shown in Fig. 3.1. The following chapters of this document will introduce the function of each module, the parameter calibration method and image quality tuning method of each module (BLC, DPC, MLSC, AWB, BNR, CCM and CLUT).

Fig. 3.1 ISP Overall structure diagram

3.3. Introduction to Each Module

The following chapters of this document will introduce the function and the image quality tuning method of each module. The functions of ISP modules are shown in Table 3.1.

Table 3.1 ISP function of each module

Module Name	Function
Crop	Function of clipping the input image.
BLC	Black level correction.
DPC	Realize the function of detecting and correcting static defect pixel and dynamic defect pixel.
CrossTalkRemoval	Correct the imbalance between Gr and Gb channels.
LSC	Provide lens shadow correction.
WDR	Wide dynamic function of multi frame synthesis.
DRC	Adjust the dynamic range of the image so that the display effect on the display device is consistent with the human visual experience.
BNR	Realize the image denoising function in Bayer domain.
Demosaic	Convert Raw image in Bayer format to RGB image.
AE	Provide automatic exposure information statistics to the software to adjust the Sensor to achieve automatic exposure function.
AWB	Provide global and regional statistical information to the software to adjust the Sensor to achieve automatic white balance function.
AF	The module outputs the statistical information related to the image definition, and the software completes the auto focusing function based on the statistical information.
CAC	Realize the function of removing purple fringing of the image, and improve the purple fringing of the image edge.
CCM	Use the 3x3 matrix to correct the color.
Gamma	Adjust the overall brightness of the image according to the gamma curve.
Dehaze	Dehaze the scene with haze to improve the contrast and clarity of the image.
CSC	Transform the RGB image into YUV image through 3x3 matrix.
YNR	Realizes the function of removing bright noise.
CNR	Provides the function of removing color noise and reducing color spots.
Sharpen	Realizes the image sharpening function and increase the image clarity.
DCI	Based on histogram equalization method to improve the overall image contrast and the details of the dark area
LDCI	Based on the method of image blocking statistics, the local contrast of the image is enhanced, and the filtering parameters can be adjusted to adjust the local range of local contrast enhancement.
CLUT	Uses 3D LUT to achieve complex color adjustment functions, including brightness adjustment and saturation adjustment.
3DNR	Removes the noise in the image by time domain filtering, keep the image details and reduce the coding bitrate.
CA	Provides saturation adjustment and thermal imaging coloring functions.
CA_Lite	Provides saturation adjustment function.
LDC	Realizes the lens deformity correction.
DIS	Digital image stabilization function.