In preprocessing, why discard top and bottom rows?

pclucas14 / lidar_generation

Code for "Deep Generative Models for LiDAR Data"

79 stars 20 forks source link

In preprocessing, why discard top and bottom rows? #17

Closed jo-jstrm closed 3 years ago

jo-jstrm commented 3 years ago

Hey again!

In parse_velo.process_velo() you discard the uppermost two and the lowest rows (lines) of the data lines = lines[2:-1].

What was the reason for that? I see that you discard more rows in utils.preprocess(), so why do not drop the aforementioned rows there and keep them during preprocessing in parse_velo.process_velo()?

I think it could have to with the discarded measurements tending to be bad in KITTI?

Best Jo

pclucas14 commented 3 years ago

Hi,

good question. I'm guessing that these rows had a lot more missing values than other rows, and were not carrying any useful info so I dropped them.

Regarding your second question, I do not recall what exactly I did in there, sorry!