RodTracker.ui.detection

TBD

class DetectorUI(ui: QWidget, image_managers: List[ImageData], *args, **kwargs)[source]

Bases: QWidget

A custom QWidget to interface with a neural network for rod detection.

Parameters:

  • ui (QWidget) – Widget containing the tab that is the GUI for the detection functionality.

  • image_managers (List[ImageData]) – List of (relevant) image data management objects providing access to loaded image datasets.

  • *args (Iterable) – Positional arguments for the QWidget superclass.

  • **kwargs (dict) – Keyword arguments for the QWidget superclass.

model

Neural network model that is used for detection.

Default is None.

Type:

ScriptModule

Signals

Slots

_progress_update(val: float, data: DataFrame, cam_id: str)[source]

Accepts progress reports of a detection process, logs and propagates them.

Parameters:

  • val (float) – Progression value of the detection process/thread \(\in [0, 1]\).

  • data (pd.DataFrame) – The DataFrame containing the detected 2D particle position data as well as the frame, color and particle numbers.

  • cam_id (str) – ID of the ImageData object for which the detection process was started.

Hint

Emits

add_color_row(color: str, amount: int, c_class: int | None = None)[source]

Add a new row to those used for the next rod detection.

Parameters:

  • color (str) – Human readable name of the class.

  • amount (int) – Number of particles that shall be detected per frame.

  • c_class (int, optional) –

    Class identifier used by the detection model.

    By default None.

autoselect_range()[source]

Not Implemented.

detected_data

Sends data of detected rods for one frame.

This signal is emitted once for every image during the detection process. The DataFrame in the payload only contains 2D position data as well as the frame, color and particle numbers.

See also

run(), rod_endpoints(), add_points()

Type:

pyqtSignal(DataFrame)

end_frame: int = 0

Last frame for detecting particles in it.

Default is 0.

Type:

int

images_loaded(num_imgs: int, id: str, path)[source]

Hook to update the available frame range for detection.

This function is intended as a slot for the data_loaded signal which in this case acts as an indicator for new image data availability. The available range of frames is updated with the information stored in the object, that emitted the signal.

Parameters:

  • num_imgs (int) – Variable to match the data_loaded signal signature. Otherwise not used.

  • id (str) – ID of the image management object, that has updated its loaded dataset. This object is then used for updating the available frames for detection of particles.

  • path (Any) – Variable to match the data_loaded signal signature. Otherwise not used.

Returns:

None

is_busy

Notifies when a background task is started/finished.

Type:

pyqtSignal(bool)

load_model()[source]

Show a file selection dialog to a user to select a particle detection model.

Lets the user select a *.pt file that should contain their desired particle detection model. The file is then loaded and the contained model set for use in the next detection(s).

Returns:

None

See also

ParticleDetection.modelling.export

number_rods: int = 1

Expected number of rods per color in one frame.

Default is 1.

Type:

int

start_detection()[source]

(Re-)Start the detection process.

Starts a detection process for each dataset loaded in the managers attribute. All frames between start_frame and end_frame are used and only the selected colors in displayed in the tab’s table will be detected. This function cannot start the detection without a loaded model.

Returns:

None

start_frame: int = 0

First frame for detecting particles in it.

Default is 0.

Type:

int

table_colors: QTableWidget = None
threshold: float = 0.5

Confidence threshold \(\in [0, 1]\) below which objects are rejected after detection.

Default is 0.5.

Type:

float

update_settings(settings: dict) → None[source]

Catches updates of the settings from a Settings class.

Checks for the keys relevant to itself and updates the corresponding attributes. Redraws itself with the new settings in place.

Parameters:

settings (dict)

Returns:

None

used_colors: List[str] = []

Colors of rods that are supposed to be detected.

Default is [].

Type:

List[str]

init_detection(ui: Ui_MainWindow, image_managers: List[ImageData])[source]

Initialize the functionality of detecting particles.

Parameters:

  • ui (Ui_MainWindow) – UI object of the main window of the application, i.e. also containing the UI tab/objects for detection tasks.

  • image_managers (List[ImageData]) – List of (relevant) image data management objects providing access to loaded image datasets. Only images from these objects will be available for particle detection.

Returns:

None | DetectorUI – Returns None, if the system requirements for particle detections are not met. Otherwise the DetectorUI object handling particle detections is returned.