SPRAD86A March   2023  – May 2024 AM62A3 , AM62A3-Q1 , AM62A7 , AM62A7-Q1 , AM67A , AM68A , AM69A

 

  1.   1
  2.   Abstract
  3.   Trademarks
  4. 1 Introduction
  5. 2 Tuning Overview
  6. 3 Hardware Requirement
  7. 4 Software Requirement
    1. 4.1 Processor SDK Linux
    2. 4.2 TI's Reference Imaging Software
    3. 4.3 ISP Tuning Tool
  8. 5 Sensor Software Integration
    1. 5.1 Overview of Image Pipeline Software Architecture
    2. 5.2 Adding Sensor Driver to SDK
    3. 5.3 Updating TIOVX Modules
      1. 5.3.1 Source Code Change
      2. 5.3.2 Rebuild Modules
    4. 5.4 Update GStreamer Plug-in for VISS
      1. 5.4.1 Update VISS Plug-in Property
      2. 5.4.2 Add Exposure Setting for 2A Algorithm
        1. 5.4.2.1 Gain
        2. 5.4.2.2 Exposure Time
        3. 5.4.2.3 Other Parameters
      3. 5.4.3 Rebuild Plug-ins
      4. 5.4.4 Verify New Sensor in GStreamer Plug-in
  9. 6 Tuning Procedure
    1. 6.1 Verify Functional Operation of Camera Capturing
    2. 6.2 Enable Camera Streaming With Initial VPAC Configuration
      1. 6.2.1 Generate Configuration Files
      2. 6.2.2 Generate DCC Binary Files
      3. 6.2.3 Stream Video With the Initial Configuration
    3. 6.3 Adjust Camera Mounting
  10. 7 Perform Basic Tuning
    1. 7.1 Launch the Tuning Tool and Create a Project
    2. 7.2 Tuning Order
    3. 7.3 Black Level Subtraction
    4. 7.4 Hardware 3A (H3A)
    5. 7.5 PCID
    6. 7.6 Auto White Balance (AWB)
      1. 7.6.1 Capture Raw Images for Different Lighting Conditions
      2. 7.6.2 Tuning AWB
    7. 7.7 Color Correction
  11. 8 Perform Fine Tuning
    1. 8.1 Edge Enhancement (EE)
    2. 8.2 Noise Filter 4 (NSF4)
  12. 9 Live Tuning
    1. 9.1 Requirements
    2. 9.2 Supported Features
      1. 9.2.1 RAW Capture
      2. 9.2.2 YUV Capture
      3. 9.2.3 Live DCC Update
      4. 9.2.4 Exposure Control
      5. 9.2.5 White Balance Control
      6. 9.2.6 Sensor Register Read/Write
  13. 10Summary
  14. 11Revision History

8.1 Edge Enhancement (EE)

The EE module parameters and LUT are automatically generated by the plug-in based on user inputs for threshold and slope.

Note:

For EE modules to take effect on AM6xA, enable ee_mode as shown in the following:

  • Open /opt/edgeai-tiovx-modules/src/tiovx_viss_module.c and find ee_mode
  • Enable edge enhancement by setting ee_mode as below:
    • obj->params.fcp[0].ee_mode = TIVX_VPAC_VISS_EE_MODE_Y8;
  • Capture a YUV image after tuning color correction. Load this image to EE plug-in of the tuning tool.
  • Set the Tuning Method to Semi-Automatic
  • Process the plug-in to switch to the EE tuning GUI. Figure 8-1 shows the EE tuning GUI and details of each control
  • Enable or disable the Edge Enhancer using the On or Off radio button in the Enable section
  • Disable the Edge Sharpener by setting the Gain to '0'
  • To see EE take effect, select one of the Image Sources in the Semi-Automatic tuning GUI and select the Apply button.

Figure 8-2 and Figure 8-3 show examples of the tuning tool output of before and after EE tuning, respectively.

EE Semi-Automatic Tuning GUI Figure 8-1 EE Semi-Automatic Tuning GUI
YUV Image Input Before EE Tuning Figure 8-2 YUV Image Input Before EE Tuning
YUV Image Output After EE Tuning Figure 8-3 YUV Image Output After EE Tuning