ntt.frames package

Submodules

ntt.frames.display module

ntt.frames.display.display_frame(frame: ndarray) → None

ntt.frames.display.display_video_as_frames(video: list | None = None) → None

ntt.frames.exif module

EXIF metadata extraction and injection for images.

Two extraction backends are provided with the same interface:

• extract_exif_pillow – uses Pillow (built-in EXIF support via _getexif)

• extract_exif_exifread – uses ExifRead (richer tag coverage, raw strings)

One injection function:

• inject_exif – embeds a metadata dict into a JPEG file (pure stdlib, no piexif)

All functions: - Accept an image file path (JPEG, TIFF for extraction; JPEG only for injection). - Raise FileNotFoundError if the file does not exist.

ntt.frames.exif.extract_exif_exifread(image_path: str) → dict

Extract EXIF metadata from an image using ExifRead.

Parameters:

image_path (str) – Path to the image file.

Returns:

Mapping of tag keys (e.g. "Image Make") to string values,

or {} if no EXIF is present.

Return type:

dict

Raises:

FileNotFoundError – If the file does not exist.

ntt.frames.exif.extract_exif_pillow(image_path: str) → dict

Extract EXIF metadata from an image using Pillow.

Parameters:

image_path (str) – Path to the image file.

Returns:

Mapping of EXIF tag names to values, or {} if no EXIF is present.

Return type:

dict

Raises:

FileNotFoundError – If the file does not exist.

ntt.frames.exif.inject_exif(image_path: str, metadata: dict) → None

Embed EXIF metadata into a JPEG file (in-place).

Supports ASCII string tags: ImageDescription, Make, Model, Software, DateTime, Artist, Copyright.

Per the TIFF spec, ASCII values whose byte count (including null terminator) fits in 4 bytes are stored inline in the IFD entry value field; longer values are stored in a separate data area and referenced by offset.

Parameters:

• image_path (str) – Path to the JPEG file to modify.

• metadata (dict) – Tag name to string value mapping. Unknown keys are silently ignored.

Raises:

• FileNotFoundError – If the file does not exist.

• ValueError – If the file is not a JPEG.

ntt.frames.frame_crop module

ntt.frames.frame_crop.crop(image: ndarray, x1: int, y1: int, x2: int, y2: int) → ndarray

Crops an image given 2 points (top left and bottom right)

Parameters:

• image (np.ndarray) – the image to crop

• x1 (int) – top left x coordinate

• y1 (int) – top left y coordinate

• x2 (int) – bottom right x coordinate

• y2 (int) – bottom right y coordinate

Raises:

ValueError – if the coordinates are out of bounds

Returns:

the cropped image

Return type:

np.ndarray

ntt.frames.frame_extraction module

ntt.frames.frame_extraction.compare_frames(video_path, frame_number)

This function compares the frames extracted by extract_frame_ffmpeg and extract_frame_opencv

Parameters:

• video_path (string) – path to the folder conataining the input video

• frame_number (int) – the number of the frame to extract.

Returns:

True if all pixels are alike between frame_opencv and frame_ffmpeg else False

Return type:

Boolean

ntt.frames.frame_extraction.extract_first_frame(video_path_in: str, video_name_in: str, frame_path_out: str, frame_name_out: str, video_URL=None) → str

This function extracts the first frame of a given video.

Parameters:

• video_path_in (string) – path to the folder conataining the input video

• video_name_in (string) – name of the input video

• frame_path_out (string) – path to the folder conataining the output frame

• frame_name_out (string) – name of the output frame

• video_URL (string, optional) – URL of the video if it is not stored locally

Returns:

full path of the output frame

Return type:

string

ntt.frames.frame_extraction.extract_frame_ffmpeg(video_path, frame_number)

Extracts a frame given its number from a video with ffmpeg

Parameters:

• video_path (string) – path to the folder conataining the input video

• frame_number (int) – the number of the frame to extract.

Returns:

the frame with number = frame_number

Return type:

np.ndarray

ntt.frames.frame_extraction.extract_frame_opencv(video_path, frame_number=1)

Extracts a frame given its number from a video with opencv

Parameters:

• video_path (string) – path to the folder conataining the input video

• frame_number (int, optional) – the number of the frame to extract. Defaults to 1.

Returns:

the frame with number = frame_number

Return type:

np.ndarray

ntt.frames.frame_extraction.extract_last_frame(video_path_in: str, video_name_in: str, frame_path_out: str, frame_name_out: str, video_URL=None) → str

Extracts the last frame of a given video.

Parameters:

• video_path_in (string) – path to the folder conataining the input video

• video_name_in (string) – name of the input video

• frame_path_out (string) – path to the folder conataining the output frame

• frame_name_out (string) – name of the output frame

• video_URL (string, optional) – URL of the video if it is not stored locally

Returns:

full path of the output frame

Return type:

string

ntt.frames.frame_extraction.extract_nth_frame(video_path_in: str, video_name_in: str, frame_path_out: str, frame_name_out: str, nth_frame=0, video_URL=None) → str

This function extracts the nth frame of a given video.

Parameters:

• video_path_in (string) – path to the folder conataining the input video

• video_name_in (string) – name of the input video

• frame_path_out (string) – path to the folder conataining the output frame

• frame_name_out (string) – name of the output frame

• nth_frame (int) – the number of the frame to be extracted

• video_URL (string, optional) – URL of the video if it is not stored locally

Returns:

full path of the output frame

Return type:

string

ntt.frames.frame_generation module

ntt.frames.frame_loading module

ntt.frames.frame_overlay module

ntt.frames.frame_overlay.overlay_n_frames(path_frames, frames, opacities, path_output_frame)

ntt.frames.frame_overlay.overlay_two_frames(path_frames, name_frame1, name_frame2, opacities, path_output_frame)

ntt.frames.io module

ntt.frames.io.read(image_path: str) → ndarray

Read an image from disk as a BGR frame.

ntt.frames.io.read_frame(image_path: str) → ndarray

Read an image from disk as a BGR frame.

ntt.frames.io.write(image_path: str, frame: ndarray) → str

Write a frame to disk and return the output path.

ntt.frames.io.write_frame(image_path: str, frame: ndarray) → str

Write a frame to disk and return the output path.

ntt.frames.metadata module

ntt.frames.metadata.extract_image_metadata(image_path: str) → dict

Return unified metadata for an image file.

ntt.frames.n_frame_extraction module

ntt.frames.n_frame_extraction.extract_frame_opencv(video_path, frame_number)

This function extracts a frame given its number from a video with opencv

Parameters:

• video_path (string) – path to the folder conataining the input video

• frame_number (int, optional) – the number of the frame to extract. Defaults to 1.

Returns:

the frame with number = frame_number

Return type:

np.ndarray

ntt.frames.n_frame_extraction.extract_n_frame(video_path_in, video_name_in, n)

This function extracts the nth frame of a given video.

Parameters:

• video_path_in (string) – path to the folder conataining the input video

• video_name_in (string) – name of the input video

• n (int) – the number of the frame to be extracted

Returns:

full path of the output frame

Return type:

string

ntt.frames.processing module

ntt.frames.processing.blur_frame(frame, kernel_size=(5, 5), region=None)

Apply Gaussian blur to a frame (NumPy array).

Parameters:

• frame (numpy.ndarray) – Input frame as a NumPy array.

• kernel_size (tuple) – Size of the Gaussian kernel for blurring.

• region (tuple) – Region of interest (x, y, width, height) to blur. If None, the entire frame will be blurred.

Returns:

Blurred frame as a NumPy array.

Return type:

numpy.ndarray

ntt.frames.processing.experiment_blur()

ntt.frames.processing.rotate(frame, angle)

ntt.frames.processing.translate_horizontally(frame, translation_rate)

ntt.frames.processing.translate_vertically(frame, translation_rate)

ntt.frames.transform module