Workflows

Room Corner Experiment

A room corner test measures fire spread across two perpendicular wall panels.

1. Set experiment_folder in config.ini and launch FlameTrack

2. Select Room Corner as experiment type

3. Load the experiment folder

4. Place 6 calibration points — 3 on the left panel, 3 on the right

5. Enter plate dimensions and click Dewarp

6. Click Find Edge — both panels are processed in parallel

7. Results are stored in the HDF5 file (see below)

Lateral Flame Spread

A lateral flame spread test measures horizontal fire spread on a single panel.

1. Load the experiment folder

2. Select Lateral Flame Spread

3. Place 4 corner points on the specimen plate

4. Enter plate dimensions and click Dewarp

5. Select flame direction and click Find Edge

6. Results are stored in the HDF5 file (see below)

Output Format

All results are saved to an HDF5 file in the processed_data/ sub-folder of the experiment directory.

Lateral Flame Spread

File: <experiment_name>_results_RCE.h5

<experiment_name>_results_RCE.h5
│  flametrack_version          str   e.g. "1.0.1"
│  flametrack_commit           str   e.g. "a3f9c12"
│  plate_width_mm              float
│  plate_height_mm             float
├── dewarped_data/
│   ├── data                   float32  (H, W, frames)   temperature per pixel per frame
│   ├── src_x                  float32  (H, W)           remap grid x
│   ├── src_y                  float32  (H, W)           remap grid y
│   └── attrs: transformation_matrix, plate_width_mm, plate_height_mm
└── edge_results/
    ├── data                   int32    (frames, H)      edge x-position per frame per row
    └── attrs: flame_direction ("left_to_right" or "right_to_left")

Room Corner

File: <experiment_name>_results_RCE.h5

<experiment_name>_results_RCE.h5
│  flametrack_version  /  flametrack_commit
│  plate_width_mm_left  /  plate_height_mm_left
│  plate_width_mm_right /  plate_height_mm_right
├── dewarped_data_left/
│   ├── data                   float32  (H, W, frames)
│   ├── src_x / src_y          float32  (H, W)
│   └── attrs: transformation_matrix, plate_width_mm, plate_height_mm
├── dewarped_data_right/
│   └── … (same structure)
├── edge_results_left/
│   └── data                   int32    (frames, H)
└── edge_results_right/
    └── data                   int32    (frames, H)

Note

edge_results/data[t, y] is the x-position of the flame edge at row y in frame t, in pixels relative to the dewarped plate. A value of 0 (L→R methods) or W (R→L methods) means no edge was detected for that row.

Room Corner experiment: calibration points on the raw IR frame (left), dewarped panels with detected edges for left and right wall (right).

Analysis view for the same experiment: flame edge position over time. Left panel (red) and right panel (blue, mirrored) with confidence band.

Reading results in Python

import h5py
import numpy as np

with h5py.File("my_experiment_results_RCE.h5", "r") as f:
    print(f.attrs["flametrack_version"])        # e.g. "1.0.1"

    edge = f["edge_results"]["data"][:]         # shape: (frames, H)
    dewarped = f["dewarped_data"]["data"]       # shape: (H, W, frames) — lazy

# edge[t, y]  →  x-position of the flame edge at row y in frame t (pixels)
# edge[t, y] == 0 or W  →  no edge detected for that row/frame