Tool/software:
Hi TI Expert,
During the deployment of our self-developed AWB (Auto White Balance) algorithm, we noticed that the TI_AWB_stab() function performs post-processing on the output results of our algorithm. We would like to understand the function and internal logic of this function. Through experiments, we found that both our self-developed AWB algorithm and TI's AWB algorithm exhibit the following identical phenomena when capturing test charts with fixed color temperature and scene in a light box:
- When the TI_AWB_stab() function is enabled and sensor_awb_pre_gain is enabled, the AWB output results can remain stable.
- When the TI_AWB_stab() function is disabled but sensor_awb_pre_gain is enabled, the AWB output results cannot stay stable.
- When both the TI_AWB_stab() function and sensor_awb_pre_gain are disabled, the AWB output results can be stable.
We understand that the TI_AWB_stab() function is designed to filter and stabilize AWB results. However, if sensor_awb_pre_gain is enabled and there are no changes in parameters such as color temperature or brightness of the image, why can the AWB results not converge when TI_AWB_stab() is disabled? What specific post-processing operations does TI_AWB_stab() perform on the AWB output results?
We hope to get your answer,thank you!

