12 FPN Correction Application

This chapter describes the GK7206 FPN (Fixed Pattern Noise) correction application example — sample_fpn. The application demonstrates the FPN correction feature, used to eliminate fixed pattern noise in sensor images.

The application source code is located in the SDK directory sample/fpn/, and serves as a reference template for developing image quality optimization applications.

1 Application Overview

1.1 Features

FPN correction: eliminates fixed pattern noise in sensor images
Interactive calibration: supports a custom number of calibration frames (1-256, powers of 2)
Offline calibration: FPN calibration in VI offline mode

1.2 Technical Parameters

Parameter	Value
`work_mode`	0 (vi offline fpn calibrate)
`Dependencies`	sensor

1.3 Directory Structure

sample/fpn/
├── Makefile              # Build script
└── sample_fpn.c          # Main program

2 Build and Deployment

2.1 Prerequisites

Before building this application, make sure the following preparations are complete:

SDK environment ready: set up the cross-compilation toolchain and SDK configuration by following SDK Build
Sensor connected: an SC465SL or compatible image sensor

2.2 Build the Application

# Enter the sample directory
cd <SDK_PATH>/sample

# Build the fpn sample
make -C fpn clean && make -C fpn

2.3 Deploy to the Board

# Mount the SD card (recommended)
mkdir -p /sd_card
mount -t vfat /dev/mmcblk1p1 /sd_card

# Transfer to the development board via SCP
scp sample/fpn/sample_fpn root@<board IP>:/sd_card/

2.4 Run the Application

# Add execute permission
chmod +x /sd_card/sample_fpn

# Enter the SD card directory
cd /sd_card

# Run (work_mode 0)
./sample_fpn 0

# Interactively enter the number of calibration frames
Enter the number of calibration frames when prompted (1-256, must be a power of 2)

# Exit
Ctrl+C

2.5 Expected Output

After a successful run, the terminal prints output similar to the following:

sample_fpn running
input number to set calibrate frame and press enter to confirm!
calibrate frame:

After the number of calibration frames is entered, FPN calibration runs and the calibration result is output.

3 Internal Execution Logic

3.1 Application Architecture

The application adopts the following architecture:

Sensor → VI → FPN calibration → calibration result saved

3.2 FPN Initialization Flow

// Step 1: system initialization
sample_comm_sys_init(&sys_config);

// Step 2: ISP initialization
sample_comm_isp_init();

// Step 3: VI start
sample_comm_vi_start(...);

// Step 4: VPSS start
sample_comm_vpss_start(...);

// Step 5: FPN initialization
xmedia_fpn_init();

// Step 6: configure FPN
xmedia_fpn_config_t fpn_config;
fpn_config.enable = XMEDIA_TRUE;
fpn_config.strength = 50;  // correction strength
xmedia_fpn_set_config(&fpn_config);

// Step 7: VENC start
sample_comm_venc_start(...);

// Step 8: bind modules
xmedia_bind_vpss_to_venc(...);

4 Key Programming Points

4.1 FPN Configuration

xmedia_fpn_config_t fpn_config;

fpn_config.enable = XMEDIA_TRUE;
fpn_config.strength = 50;  // correction strength (0~100)
fpn_config.mode = XMEDIA_FPN_MODE_AUTO;  // auto mode

xmedia_fpn_set_config(&fpn_config);

4.2 FPN Enable/Disable

// Enable FPN
xmedia_fpn_enable(XMEDIA_TRUE);

// Disable FPN
xmedia_fpn_enable(XMEDIA_FALSE);

5 Troubleshooting

Problem	Possible Cause	Solution
FPN effect not obvious	Correction strength set too low	Adjust the strength parameter to 50~80
Picture too dark or too bright	Incorrect FPN configuration	Check the FPN configuration parameters
Initialization fails	Sensor does not support FPN	Check the sensor model