HI,
What is the meaning of the difference of four figure?
histogram of C2M distances
histogram of C2M distances
- Attachments
-
- Histogramc2m for all.jpg (151.55 KiB) Viewed 1334 times
Re: histogram of C2M distances
1) the distance distribution is not the same for each case ;)
2) The average is always close to zero (I guess a registration process was applied?)
3) the std. deviation is the most meaningful (it gives an information on well each part matches - the narrower the better
4) you can forget about the 95% value here as this is a signed distribution, so what you would like is to look at the absolute distance that encompass 95% of the values. It should be around 2 times the standard distribution (well, if your distribution was actually 'Normal', which is not exactly the case sadly). Or you can try to look at the negative value that catches 2.5% of the values, and the positive value that catches 97.5% of values, and hope there absolute value is more or less the same...
2) The average is always close to zero (I guess a registration process was applied?)
3) the std. deviation is the most meaningful (it gives an information on well each part matches - the narrower the better
4) you can forget about the 95% value here as this is a signed distribution, so what you would like is to look at the absolute distance that encompass 95% of the values. It should be around 2 times the standard distribution (well, if your distribution was actually 'Normal', which is not exactly the case sadly). Or you can try to look at the negative value that catches 2.5% of the values, and the positive value that catches 97.5% of values, and hope there absolute value is more or less the same...
Daniel, CloudCompare admin