openflexure_microscope_server.things.camera.picamera_tuning_file_utils

module documentation
(source)

Functions for loading, adjusting, or reading from the Picamera2 tuning file.

The functions that edit the tuning files return a new dictionary that is updated.

Class LensShadingModel A Pydantic model holding the lens shading tables.
Exception TuningFileError Raised if the tuning file cannot be loaded for any reason.
Function ce_enable_is_static Whether the ce_enable flag is disabled.
Function copy_algo_from_other_tuning Return a copy of tuning_file with an algorithm copied from another file.
Function find_tuning_algo Return the parameters for the named algorithm in the given camera tuning dict.
Function flatten_lst Flaten the len shading table ro an array of ones.
Function geq_is_static Whether the green equalisation is set to static.
Function get_ccm Get a copy of the the rpi.ccm section of a camera tuning dict.
Function get_colour_gains_from_lst Get the colour gains that are needed from the lens shading tables.
Function get_lst Return the lens shading as a LenSading Base Model.
Function load_default_tuning Load the default tuning file for the camera.
Function lst_calibrated Whether the lens shading table is calibrated.
Function set_ccm Update the rpi.ccm section of a camera tuning dict set the colour correction matrix.
Function set_ce_to_disabled Set ce_enable in rpi.contrast to zero to disable adaptive contrast enhancement.
Function set_lst Update the rpi.alsc section of with new lens shading tables.
Function set_static_geq Update the rpi.geq section of a camera tuning dict.
Constant CALIBRATED_COLOUR_TEMP Undocumented
Constant DEFAULT_COLOUR_TEMP Undocumented
Constant THIS_DIR Undocumented
Function _as_flat_rounded_list Flatten array, round, and then convert to list.
Function _index_of_algorithm Find the index of an algorithm's section in the tuning file.
def ce_enable_is_static(tuning: dict) -> bool: (source)

Whether the ce_enable flag is disabled.

def copy_algo_from_other_tuning(algo: str, *, base_tuning_file: dict, copy_from: dict) -> dict: (source)

Return a copy of tuning_file with an algorithm copied from another file.

Tuning dict arguments are keyword only for clarity.

Parameters
algo:strThe algorithm to copy. Eg rpi.alsc for lens shading correction.
base_tuning_file:dictThe tuning file to copy.
copy_from:dictThe tuning file to copy the algorithm from section from.
Returns
dictA deep copy of base_tuning_file with the specified algorithm copied in from the other tuning file.
def find_tuning_algo(tuning: dict[str, dict], name: str) -> dict[str, Any]: (source)

Return the parameters for the named algorithm in the given camera tuning dict.

This is the same methodolgy used in the PiCamera2 library but is provided here so it can be tested independently of installing picamera2

Parameters
tuning:dict[str, dict]The camera tuning dictionary
name:strThe key for the algorithm in the tuning file
Returns
dict[str, Any]The algorithm from the tuning dictionary. Editing this will edit the original dictionary.
def flatten_lst(tuning: dict, keep_luminance: bool = False) -> dict: (source)

Flaten the len shading table ro an array of ones.

Parameters
tuning:dictThe current tuning dictionary.
keep_luminance:boolSet to True to only flatten the cr and cb tables.
Returns
dictAn updated tuning dict.
def geq_is_static(tuning: dict) -> bool: (source)

Whether the green equalisation is set to static.

def get_ccm(tuning: dict): (source)

Get a copy of the the rpi.ccm section of a camera tuning dict.

def get_colour_gains_from_lst(tuning: dict) -> tuple[float, float]: (source)

Get the colour gains that are needed from the lens shading tables.

The lens shading tables are calculated to create a white balanced image, but the ISP normalises by the minimum Cr and Cb value. So these need to be set as colour gains.

def get_lst(tuning: dict) -> LensShadingModel: (source)

Return the lens shading as a LenSading Base Model.

def load_default_tuning(sensor_model: str) -> dict: (source)

Load the default tuning file for the camera.

This will load the tuning file based on the specified sensor model.

def lst_calibrated(tuning: dict) -> bool: (source)

Whether the lens shading table is calibrated.

This checks whether the lens shading table is has a colour temperature of 5000. As this is what we set on calibration. Our tuning file sets a temperature of 1234.

def set_ccm(tuning: dict, col_corr_matrix: list) -> dict: (source)

Update the rpi.ccm section of a camera tuning dict set the colour correction matrix.

Parameters
tuning:dictThe current tuning dict
col_corr_matrix:listThe colour correction matrix to set
Returns
dictan updated tuning dict with the new colour correction matrix.
def set_ce_to_disabled(tuning: dict) -> dict: (source)

Set ce_enable in rpi.contrast to zero to disable adaptive contrast enhancement.

Parameters
tuning:dictThe raspberry pi camera tuning file.
Returns
dictA deepcopy of the input file with ce_enable set to 0.
def set_lst(tuning: dict, *, luminance: np.ndarray | None, cr: np.ndarray | None, cb: np.ndarray | None, colour_temp: int) -> dict: (source)

Update the rpi.alsc section of with new lens shading tables.

Only one set of tables is set so that the camera pipeline does not adaptively switch between lens shading tables based on its estimation of colour temperature. Also n_iter is set to 0 so that the pipeline doesn't perform an iterative adaption of our table to try to "correct" if interprets the image as having different types of lighting.

Parameters
tuning:dictThe current tuning file.
luminance:np.ndarray | NoneThe table of luminance values, as (12, 16) numpy array. Or None to leave unchanged.
cr:np.ndarray | NoneThe table of Cr values, as (12, 16) numpy array. Or None to leave unchanged.
cb:np.ndarray | NoneThe table of Cb values, as (12, 16) numpy array. Or None to leave unchanged.
colour_temp:intThe colour temperature to set. On calibration this should be set to 5000. Set a different value for the PiCamera Thing to report that the lens shading is not calibrated.
Returns
dictan updated tuning dict with the new lens shading tables.
def set_static_geq(tuning: dict, offset: int = 65535) -> dict: (source)

Update the rpi.geq section of a camera tuning dict.

Parameters
tuning:dictthe Raspberry Pi camera tuning dictionary
offset:intThe desired green equalisation offset. Default 65535. The default is the maximum allowed value. This means the brightness will always be below the threshold where averaging is used. This is default as we always need the green equalisation to averages the green pixels in the red and blue rows due to the chief ray angle compensation issue when the the stock lens is replaced by an objective.
Returns
dictAn updated tuning dictionary
CALIBRATED_COLOUR_TEMP: int = (source)

Undocumented

Value
5000
DEFAULT_COLOUR_TEMP: int = (source)

Undocumented

Value
1234
THIS_DIR = (source)

Undocumented

Value
os.path.dirname(os.path.abspath(__file__))
def _as_flat_rounded_list(array: np.ndarray, round_to: int = 3) -> list[float]: (source)

Flatten array, round, and then convert to list.

def _index_of_algorithm(algorithms: list[dict], algorithm: str) -> int: (source)

Find the index of an algorithm's section in the tuning file.