Project

General

Profile

Crash "Filters clusters by the distance of the bounding box to a given point"

Added by Georgi Louis about 3 years ago

Hallo CompuTree community,

I´m using Computree 4.0.759. With the sample data and jans script everything is working fine.

Applying those steps to my own data (.xyz/ .xyb), computree crashes always at the step: Filters clusters by the distance of the bounding box to a given point (log file atteched).

Randomly the warning of a “georeferenced pointcloud” is mentioned. But not always with the same cloud. Hence I shifted my cloud.

But computree is still crashing.

Any idea?

Thanks a lot and best regards

Louis

Nuage de points, Fichiers ASCII (paramétrable) (1) - Start computing at 14:05:45
Nuage de points, Fichiers ASCII (paramétrable) (1) - Computing completed, elapsed time: 0h:1m:21s:848ms
TK_TranslateCloud (2) - Start computing at 14:07:07
TK_TranslateCloud (2) - Computing completed, elapsed time: 0h:0m:3s:427ms
ONF_StepClassifyGround (3) - Start computing at 14:07:10
The 1 input scene contains 26920399 points.
DTM grid to create: 520 rows and 556 columns
Zmin grid created
Filtering on density achieved
Neighbourhood consistency test achieved
Scene 1: soil (15610788 points) and vegetation (11309611 points) scenes creation achieved
ONF_StepClassifyGround (3) - Computing completed, elapsed time: 0h:0m:7s:960ms
ST_StepFilterGroundPoints (4) - Start computing at 14:07:18
The input cloud was downscaled to a cloud with 0.15% points remaining.
The input cloud was downscaled to a cloud with 7.05% points remaining.
The normal orientation of the ground points is (-0.02;0.07;1.00).
ST_StepFilterGroundPoints (4) - Computing completed, elapsed time: 0h:0m:47s:520ms
ST_StepComputeDTM (5) - Start computing at 14:08:06
The input cloud was downscaled to a cloud with 0.75% points remaining.
ST_StepComputeDTM (5) - Computing completed, elapsed time: 0h:0m:10s:282ms
ST_StepExtactSliceAboveDTM (6) - Start computing at 14:08:16
La scène d’entrée comporte 11309611 points.
La scène extraite comporte 3144770 points.
ST_StepExtactSliceAboveDTM (6) - Computing completed, elapsed time: 0h:0m:18s:663ms
ST_StepFilterStems (7) - Start computing at 14:08:35
The input cloud was downscaled to a cloud with 0.15% points remaining.
The input cloud was downscaled to a cloud with 7.05% points remaining.
The normal orientation of the ground points is (-0.02;0.07;1.00).
The input cloud was downscaled to a cloud with 23.81% points remaining.
ST_StepFilterStems (7) - Computing completed, elapsed time: 0h:0m:19s:455ms
ST_StepExtractLargestCluster (8) - Start computing at 14:08:54
ST_StepExtractLargestCluster (8) - Computing completed, elapsed time: 0h:0m:13s:759ms
ST_StepExtactSliceAboveDTM (9) - Start computing at 14:09:08
La scène d’entrée comporte 11309611 points.
La scène extraite comporte 8085867 points.
ST_StepExtactSliceAboveDTM (9) - Computing completed, elapsed time: 0h:0m:19s:156ms
ST_StepMergeClouds (10) - Start computing at 14:09:27
ST_StepMergeClouds (10) - Computing completed, elapsed time: 0h:0m:0s:807ms
ST_StepVoxelGridFilter (11) - Start computing at 14:09:28
The input cloud was downscaled to a cloud with 99.91% points remaining.
ST_StepVoxelGridFilter (11) - Computing completed, elapsed time: 0h:0m:5s:22ms
ST_StepClearSky (12) - Start computing at 14:09:33
ST_StepClearSky (12) - Computing completed, elapsed time: 0h:0m:19s:392ms
ST_StepExtactSliceAboveDTM (13) - Start computing at 14:09:52
La scène d’entrée comporte 9656594 points.
La scène extraite comporte 397214 points.
ST_StepExtactSliceAboveDTM (13) - Computing completed, elapsed time: 0h:0m:15s:180ms
ST_StepEuclideanClustering (14) - Start computing at 14:10:07
ST_StepEuclideanClustering (14) - Computing completed, elapsed time: 0h:0m:14s:650ms
ST_StepSegmentationAll (15) - Start computing at 14:10:22
ST_StepSegmentationAll (15) - Computing completed, elapsed time: 0h:0m:50s:757ms
ST_StepFilterClusters (16) - Start computing at 14:11:13
Number of clouds before filtering : 170
Number of clouds rejected : 9
ST_StepFilterClusters (16) - Computing completed, elapsed time: 0h:0m:0s:2ms
ST_StepFilterClustersByDistance (17) - Start computing at 14:11:13


Replies (1)

RE: Crash "Filters clusters by the distance of the bounding box to a given point" - Added by Hackenberg Jan almost 3 years ago

Hi

first of all yes it is good that you center your cloud. It is difficult to tell why the pipeline is crashing. I quickly checked the code of the crashing step. In case it gets an empty cloud as input, it will cause a crash (division by zero).
a) you should not have an empty cloud, because those should be filtered in the stem before (16) before from what I see. I do not know your parameters, but you remove 9 clouds which presumingly have too small bounding boxes.

b) You should try out by placing the step “filters clusters by point number” before. Please report if this helps.

c) if it does not help I need to get access to one cloud where this crash occurs. Please do not sent a cloud by mail and contact me before. Make sure you tried out b) first.

I think in general too small or empty clouds can cause havoc with my plugin. You can also try out to export all segmented trees remaing after step 16 and view them one by one. Only possible in a small plot. But maybe you see something there.

    (1-1/1)