from collections import OrderedDict
from glob import glob
import os
from typing import Tuple
import pandas as pd
from perceptionmetrics.datasets import segmentation as segmentation_dataset
import perceptionmetrics.utils.io as uio
# Default split presented in the paper
DEFAULT_SPLIT = {
"creek": "test",
"park-1": "test",
"park-2": "train",
"park-8": "val",
"trail": "train",
"trail-3": "train",
"trail-4": "train",
"trail-5": "val",
"trail-6": "train",
"trail-7": "test",
"trail-9": "train",
"trail-10": "train",
"trail-11": "train",
"trail-12": "train",
"trail-13": "test",
"trail-14": "train",
"trail-15": "train",
"village": "train",
}
[docs]
def build_dataset(
data_dir: str,
labels_dir: str,
ontology_fname: str,
split_sequences: dict,
) -> Tuple[dict, dict]:
"""Build dataset and ontology dictionaries
:param data_dir: Directory containing data
:type data_dir: str
:param labels_dir: Directory containing labels (in RGB format)
:type labels_dir: str
:param ontology_fname: text file containing the dataset ontology (RUGD_annotation-colormap.txt)
:type ontology_fname: str
:param split_sequences: Dictionary containing the split sequences for train, val, and test
:type split_sequences: dict
:return: Dataset and onotology
:rtype: Tuple[dict, dict]
"""
# Check that provided paths exist and ensure they are absolute
data_dir = os.path.abspath(data_dir)
labels_dir = os.path.abspath(labels_dir)
assert os.path.isdir(data_dir), "Images directory not found"
assert os.path.isdir(labels_dir), "Labels directory not found"
# Load and adapt ontology
assert os.path.isfile(ontology_fname), "Ontology file not found"
original_ontology = uio.read_txt(ontology_fname)
try:
ontology = {}
for class_data in original_ontology:
class_idx, class_name, r, g, b = class_data.split(" ")
ontology[class_name] = {
"idx": int(class_idx),
"rgb": (int(r), int(g), int(b)),
}
except Exception as e:
print(f"Error loading ontology: {e}")
raise
# Build dataset as ordered python dictionary
dataset = OrderedDict()
for data_fname in glob(os.path.join(data_dir, "*/*.png")):
label_fname = data_fname.replace(data_dir, labels_dir)
assert os.path.isfile(label_fname), f"Label file not found: {label_fname}"
sample_name, _ = os.path.splitext(os.path.basename(data_fname))
scene_name = sample_name.split("_")[0]
split = split_sequences[scene_name]
dataset[sample_name] = (data_fname, label_fname, split)
return dataset, ontology
[docs]
class RUGDImageSegmentationDataset(segmentation_dataset.ImageSegmentationDataset):
"""Specific class for RUGD-styled image segmentation dataset.
:param images_dir: Directory containing images
:type images_dir: str
:param labels_dir: Directory containing labels (in RGB format)
:type labels_dir: str
:param ontology_fname: text file containing the dataset ontology (RUGD_annotation-colormap.txt)
:type ontology_fname: str
:param split_sequences: Dictionary containing the split sequences for train, val, and test, defaults to DEFAULT_SPLIT
:type split_sequences: dict, optional
.. seealso::
`User guide and format examples <https://jderobot.github.io/PerceptionMetrics/datasets/#rugd>`__.
"""
def __init__(
self,
images_dir: str,
labels_dir: str,
ontology_fname: str,
split_sequences: dict = DEFAULT_SPLIT,
):
dataset, ontology = build_dataset(
images_dir,
labels_dir,
ontology_fname,
split_sequences,
)
# Convert to Pandas
cols = ["image", "label", "split"]
dataset = pd.DataFrame.from_dict(dataset, orient="index", columns=cols)
# Report results
print(f"Samples retrieved: {len(dataset)}")
# Initialize parent class
super().__init__(dataset, images_dir, ontology, is_label_rgb=True)