Application on point cloud

Load modules

import numpy
import pandas

import openalea.phenomenal.display.notebook as phm_display_notebook
import openalea.phenomenal.object as phm_obj
import openalea.phenomenal.segmentation as phm_seg

Load cloud point cloud

First we define a function “read_from_xyz” to load our point cloud data

def read_from_xyz(filename):

    xyz_position = list()
    with open(filename, 'r') as f:
        for line in f:
            values = [float(v) for v in line.split()[:3]] # load just position not color
            xyz_position.append(tuple(values))
    f.close()

    return numpy.array(xyz_position)

Maize point cloud was found on : CharlieLeee/Maize-plant-point-cloud-dataset and cleaned mannually with CloudCompare software. 3D reconstruction is pretty bad, leaves are very noised and seems cutted.

file_path = "maize_point_cloud.pts"
xyz_position = read_from_xyz(file_path)
phm_display_notebook.show_point_cloud(xyz_position, size=0.1)

Convert point cloud to voxel grid

Once loaded, we attribute for each point a fictive voxel size to simulate the data like a voxel grid. More the fictive voxel size is small more the 3D voxel representation is accurate. After we normalize the pointcloud into a grid.

fictive_voxel_size = 0.05

voxel_grid = phm_obj.VoxelGrid(numpy.array(xyz_position), fictive_voxel_size)
voxel_grid = phm_obj.VoxelGrid.from_image_3d(
    voxel_grid.to_image_3d(),
    voxels_value=1,
    voxels_size=1, # must be integer
    world_coordinate=(0.0, 0.0, 0.0))

print("Shape Image 3D:", voxel_grid.to_image_3d().shape)

phm_display_notebook.show_voxel_grid(voxel_grid, size=1)

Shape Image 3D: (134, 212, 96)

graph = phm_seg.graph_from_voxel_grid(voxel_grid, connect_all_point=False)
src_node = tuple(max(graph.nodes(), key=lambda d: d[1]))
voxel_skeleton = phm_seg.skeletonize(voxel_grid, graph, src_node=src_node)

Skeletonization

phm_display_notebook.show_skeleton(voxel_skeleton, with_voxel=True, size=1.0)

Maize Segmentation

vms = phm_seg.maize_segmentation(voxel_skeleton, graph, stem_strategy="longest")

phm_display_notebook.show_segmentation(vms, size=1)

Maize Analysis

vmsi = phm_seg.maize_analysis(vms)
phm_display_notebook.show_segmentation(vmsi, size=1)

Take a look, of what kind of data is extract. (pm = phenomenal_mearsurement)

df = pandas.DataFrame([vo.info for vo in vmsi.voxel_organs]  + [vmsi.info])
df

	pm_label	pm_sub_label	pm_voxels_volume	pm_position_tip	pm_position_base	pm_z_tip	pm_z_base	pm_length	pm_width_max	pm_width_mean	...	pm_azimuth_angle	pm_inclination_angle	pm_full_length	pm_length_with_speudo_stem	pm_insertion_angle_vector	pm_insertion_angle	pm_length_speudo_stem	pm_leaf_number	pm_z_base_voxel	pm_number_of_leaf
0	unknown	NaN	186	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1	stem	NaN	2337	(71.0, 85.0, 44.0)	(81.0, 211.0, 61.0)	44.0	61.0	136.316379	11.401754	7.013952	...	-94.317102	0.775932	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
2	growing_leaf	NaN	6907	(2.0, 7.0, 57.0)	(71.0, 84.0, 44.0)	57.0	44.0	119.181579	26.907248	14.662080	...	-122.668166	0.893494	256.497959	118.181579	(-2.6956521739130435, -12.0, 0.0)	11.526682	-1.000000	7.0	NaN	NaN
3	growing_leaf	NaN	3556	(71.0, 21.0, 9.0)	(75.0, 56.0, 31.0)	9.0	31.0	46.212474	12.688578	9.697847	...	-91.091216	3.433875	215.545905	78.183037	(0.0, -5.0, -4.444444444444445)	33.868735	31.970563	8.0	NaN	NaN
4	growing_leaf	NaN	4158	(58.0, 14.0, 47.0)	(63.0, 34.0, 48.0)	47.0	48.0	22.388905	9.273618	5.549971	...	-100.784298	4.636406	214.290342	75.973963	(-0.6, -3.0, 0.0)	10.596425	53.585057	9.0	NaN	NaN
5	mature_leaf	NaN	567	(77.0, 182.0, 9.0)	(80.0, 184.0, 52.0)	9.0	52.0	64.193966	7.615773	4.439786	...	-120.424395	2.922740	95.239725	NaN	(-3.4166666666666665, -6.083333333333333, -6.5)	41.293851	NaN	5.0	53.0	NaN
6	mature_leaf	NaN	1329	(128.0, 147.0, 40.0)	(85.0, 175.0, 53.0)	40.0	53.0	60.537111	15.033296	7.700353	...	-41.228776	2.291478	101.122169	NaN	(4.090909090909091, -6.0, -1.4545454545454546)	38.177025	NaN	6.0	52.0	NaN
7	mature_leaf	NaN	1484	(24.0, 120.0, 56.0)	(71.0, 152.0, 51.0)	56.0	51.0	69.504274	11.874342	6.881522	...	-142.300788	1.645193	136.113911	NaN	(-7.0, -6.384615384615385, -1.0769230769230769)	42.978219	NaN	4.0	47.0	NaN
8	mature_leaf	NaN	1699	(122.0, 85.0, 31.0)	(82.0, 134.0, 49.0)	31.0	49.0	73.632759	13.379088	8.227598	...	-58.572101	1.979836	155.921159	NaN	(1.4615384615384615, -7.0, -1.0)	15.953872	NaN	2.0	38.0	NaN
9	mature_leaf	NaN	3380	(0.0, 55.0, 86.0)	(71.0, 106.0, 49.0)	86.0	49.0	108.748857	17.029386	9.709537	...	-136.528551	0.891515	222.090545	NaN	(-8.0, -10.0, 1.105263157894737)	36.610113	NaN	3.0	47.0	NaN
10	mature_leaf	NaN	3009	(115.0, 42.0, 26.0)	(79.0, 101.0, 45.0)	26.0	45.0	95.271833	21.954498	10.111130	...	-67.995238	1.419182	213.170599	NaN	(2.2777777777777777, -9.5, -2.5)	18.577649	NaN	1.0	23.0	NaN
11	plant	NaN	22575	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	9.0

12 rows × 25 columns