How can you tell?
Surely the point you're making is that the real reason for pulling the person over is not the one given, how then can you establish it's "racial profiling".
I've been pulled over (UK) by police for having a stop light out, reasonable as it's a traffic violation. They asked the usual sort of questions and I got a "producer" (notice to produce various documents at a police station in the near future). How could you establish that this was because the officer was acting racist towards me?
Telling the race of a person from behind when following their car seems quite hard.
Statistics could show that more of a particular race were pulled over but that race could constitute a greater proportion of the local poor who being less able to keep their vehicles in good order are more likely to be stopped. That would create an apparent bias were there was none.
It's why young African Americans are more likely to be arrested for minor drug offenses, despite offending less then the average young white person. (see http://www.washingtonpost.com/blogs/wonkblog/wp/2013/06/04/t... ) And indeed if you look at the ACLU Report on disparities between whites and blacks for minor marijuana offenses, page 178 has Waller County, where this happened, as one of the ones with disparities in arrests above the national average. Note: It's a PDF: https://www.aclu.org/files/assets/aclu-thewaronmarijuana-rel...