Part Number: TDA4VE-Q1
HiTI's engineer:
Excuse me, which path is the source code of H3A module? At present, we want to try to optimize the statistical accuracy of H3A module and find the source code to modify it.
This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
Part Number: TDA4VE-Q1
HiTI's engineer:
Excuse me, which path is the source code of H3A module? At present, we want to try to optimize the statistical accuracy of H3A module and find the source code to modify it.
Because H3A can only count 10bits, the accuracy is not high. At present, we want to realize the function of calculating bv and ev, and we need to obtain the raw data input to H3A after pwl decompression. Can you give us some instructions on how to realize this function? thank you
Hi,
I don't quite understand what you are asking for.
Because H3A can only count 10bits, the accuracy is not high.
Yes, this is H/W limitation.
At present, we want to realize the function of calculating bv and ev
I suppose you mean deciding sensor exposure.
What do "bv" (?) and "ev" (exposure value?) stand for?
we need to obtain the raw data input to H3A after pwl decompression
You won't have access to this on H/W.
What do you need this for?
Can you give us some instructions on how to realize this function?
The function of deciding "bv" and "ev"?
We want to calculate the brightness of the current environment (not the brightness of the current picture). We need to obtain the original raw statistical data, exposure time and gain to convert the brightness of the current environment, so as to ensure the accuracy, so that we can judge different scenes later (such as tunnel scenes, sunny scenes, cloudy scenes and night scenes). Different parameters can be used in different scenes to ensure the accuracy of the effect.
At present, the accuracy of statistical data of H3A can't meet the accuracy of environmental brightness that we want to calculate at present.
We need to obtain the original raw statistical data,
What kind of statistical data are you looking for?
It looks like you will have to use the raw sensor image to do the statistics in DPS/ARM.
Another possibility is to use sensor embedded data if it supports the kind of statistic you need.
We need raw data after pwl decompression
This is internal to the ISP pipeline -- not accessible out of ISP.
We want to get the raw data after pwl decompression to calculate the brightness
You will have to do decompanding in DSP using the raw sensor image before collecting statistics.