Detecting outliers with RANSAC

2 次查看（过去 30 天）

显示更早的评论

Peter Larsen 2019-5-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/462230-detecting-outliers-with-ransac

I have a matrix out (150x2), which looks like this:

out=[ 12 5.708

12 -1.765

12 -1.760

12 -1.755

13 0.015

13 -0.027

13 0.002

15 0.0568

15 0.098]

The first column are point ID, the second column are residuals. The outlier limit for the residuals are +-1.7. That gives a problem with point ID 12, because of the large deviation of 5.708, while the other residuals of point ID 12 are close to each other. Is there a way to use RANSAC for each point ID.

I have tried to use the build in RANSAC function:

[model,inlierIdx] = ransac(data,fitFcn,distFcn,sampleSize,maxDistance)

data=out

maxDistance=1.7

But I can't make it work for each point ID.