"""Contains everything that can help automate the cuts in MoviePy.""" from collections import defaultdict import numpy as np from moviepy.decorators import convert_parameter_to_seconds, use_clip_fps_by_default @use_clip_fps_by_default @convert_parameter_to_seconds(["start_time"]) def find_video_period(clip, fps=None, start_time=0.3): """Find the period of a video based on frames correlation. Parameters ---------- clip : moviepy.Clip.Clip Clip for which the video period will be computed. fps : int, optional Number of frames per second used computing the period. Higher values will produce more accurate periods, but the execution time will be longer. start_time : float, optional First timeframe used to calculate the period of the clip. Examples -------- .. code:: python from moviepy import * from moviepy.video.tools.cuts import find_video_period clip = VideoFileClip("media/chaplin.mp4").subclipped(0, 1).loop(2) round(videotools.find_video_period(clip, fps=80), 6) 1 """ def frame(t): return clip.get_frame(t).flatten() timings = np.arange(start_time, clip.duration, 1 / fps)[1:] ref = frame(0) corrs = [np.corrcoef(ref, frame(t))[0, 1] for t in timings] return timings[np.argmax(corrs)] class FramesMatch: """Frames match inside a set of frames. Parameters ---------- start_time : float Starting time. end_time : float End time. min_distance : float Lower bound on the distance between the first and last frames max_distance : float Upper bound on the distance between the first and last frames """ def __init__(self, start_time, end_time, min_distance, max_distance): self.start_time = start_time self.end_time = end_time self.min_distance = min_distance self.max_distance = max_distance self.time_span = end_time - start_time def __str__(self): # pragma: no cover return "(%.04f, %.04f, %.04f, %.04f)" % ( self.start_time, self.end_time, self.min_distance, self.max_distance, ) def __repr__(self): # pragma: no cover return self.__str__() def __iter__(self): # pragma: no cover return iter( (self.start_time, self.end_time, self.min_distance, self.max_distance) ) def __eq__(self, other): return ( other.start_time == self.start_time and other.end_time == self.end_time and other.min_distance == self.min_distance and other.max_distance == self.max_distance ) class FramesMatches(list): """Frames matches inside a set of frames. You can instantiate it passing a list of FramesMatch objects or using the class methods ``load`` and ``from_clip``. Parameters ---------- lst : list Iterable of FramesMatch objects. """ def __init__(self, lst): list.__init__(self, sorted(lst, key=lambda e: e.max_distance)) def best(self, n=1, percent=None): """Returns a new instance of FramesMatches object or a FramesMatch from the current class instance given different conditions. By default returns the first FramesMatch that the current instance stores. Parameters ---------- n : int, optional Number of matches to retrieve from the current FramesMatches object. Only has effect when ``percent=None``. percent : float, optional Percent of the current match to retrieve. Returns ------- FramesMatch or FramesMatches : If the number of matches to retrieve is greater than 1 returns a FramesMatches object, otherwise a FramesMatch. """ if percent is not None: n = len(self) * percent / 100 return self[0] if n == 1 else FramesMatches(self[: int(n)]) def filter(self, condition): """Return a FramesMatches object obtained by filtering out the FramesMatch which do not satistify a condition. Parameters ---------- condition : func Function which takes a FrameMatch object as parameter and returns a bool. Examples -------- .. code:: python # Only keep the matches corresponding to (> 1 second) sequences. new_matches = matches.filter(lambda match: match.time_span > 1) """ return FramesMatches(filter(condition, self)) def save(self, filename): """Save a FramesMatches object to a file. Parameters ---------- filename : str Path to the file in which will be dumped the FramesMatches object data. """ np.savetxt( filename, np.array([np.array(list(e)) for e in self]), fmt="%.03f", delimiter="\t", ) @staticmethod def load(filename): """Load a FramesMatches object from a file. Parameters ---------- filename : str Path to the file to use loading a FramesMatches object. Examples -------- >>> matching_frames = FramesMatches.load("somefile") """ arr = np.loadtxt(filename) mfs = [FramesMatch(*e) for e in arr] return FramesMatches(mfs) @staticmethod def from_clip(clip, distance_threshold, max_duration, fps=None, logger="bar"): """Finds all the frames that look alike in a clip, for instance to make a looping GIF. Parameters ---------- clip : moviepy.video.VideoClip.VideoClip A MoviePy video clip. distance_threshold : float Distance above which a match is rejected. max_duration : float Maximal duration (in seconds) between two matching frames. fps : int, optional Frames per second (default will be ``clip.fps``). logger : str, optional Either ``"bar"`` for progress bar or ``None`` or any Proglog logger. Returns ------- FramesMatches All pairs of frames with ``end_time - start_time < max_duration`` and whose distance is under ``distance_threshold``. Examples -------- We find all matching frames in a given video and turn the best match with a duration of 1.5 seconds or more into a GIF: .. code:: python from moviepy import VideoFileClip from moviepy.video.tools.cuts import FramesMatches clip = VideoFileClip("foo.mp4").resize(width=200) matches = FramesMatches.from_clip( clip, distance_threshold=10, max_duration=3, # will take time ) best = matches.filter(lambda m: m.time_span > 1.5).best() clip.subclipped(best.start_time, best.end_time).write_gif("foo.gif") """ N_pixels = clip.w * clip.h * 3 def dot_product(F1, F2): return (F1 * F2).sum() / N_pixels frame_dict = {} # will store the frames and their mutual distances def distance(t1, t2): uv = dot_product(frame_dict[t1]["frame"], frame_dict[t2]["frame"]) u, v = frame_dict[t1]["|F|sq"], frame_dict[t2]["|F|sq"] return np.sqrt(u + v - 2 * uv) matching_frames = [] # the final result. for t, frame in clip.iter_frames(with_times=True, logger=logger): flat_frame = 1.0 * frame.flatten() F_norm_sq = dot_product(flat_frame, flat_frame) F_norm = np.sqrt(F_norm_sq) for t2 in list(frame_dict.keys()): # forget old frames, add 't' to the others frames # check for early rejections based on differing norms if (t - t2) > max_duration: frame_dict.pop(t2) else: frame_dict[t2][t] = { "min": abs(frame_dict[t2]["|F|"] - F_norm), "max": frame_dict[t2]["|F|"] + F_norm, } frame_dict[t2][t]["rejected"] = ( frame_dict[t2][t]["min"] > distance_threshold ) t_F = sorted(frame_dict.keys()) frame_dict[t] = {"frame": flat_frame, "|F|sq": F_norm_sq, "|F|": F_norm} for i, t2 in enumerate(t_F): # Compare F(t) to all the previous frames if frame_dict[t2][t]["rejected"]: continue dist = distance(t, t2) frame_dict[t2][t]["min"] = frame_dict[t2][t]["max"] = dist frame_dict[t2][t]["rejected"] = dist >= distance_threshold for t3 in t_F[i + 1 :]: # For all the next times t3, use d(F(t), F(end_time)) to # update the bounds on d(F(t), F(t3)). See if you can # conclude on whether F(t) and F(t3) match. t3t, t2t3 = frame_dict[t3][t], frame_dict[t2][t3] t3t["max"] = min(t3t["max"], dist + t2t3["max"]) t3t["min"] = max(t3t["min"], dist - t2t3["max"], t2t3["min"] - dist) if t3t["min"] > distance_threshold: t3t["rejected"] = True # Store all the good matches (end_time,t) matching_frames += [ (t1, t, frame_dict[t1][t]["min"], frame_dict[t1][t]["max"]) for t1 in frame_dict if (t1 != t) and not frame_dict[t1][t]["rejected"] ] return FramesMatches([FramesMatch(*e) for e in matching_frames]) def select_scenes( self, match_threshold, min_time_span, nomatch_threshold=None, time_distance=0 ): """Select the scenes at which a video clip can be reproduced as the smoothest possible way, mainly oriented for the creation of GIF images. Parameters ---------- match_threshold : float Maximum distance possible between frames. The smaller, the better-looping the GIFs are. min_time_span : float Minimum duration for a scene. Only matches with a duration longer than the value passed to this parameters will be extracted. nomatch_threshold : float, optional Minimum distance possible between frames. If is ``None``, then it is chosen equal to ``match_threshold``. time_distance : float, optional Minimum time offset possible between matches. Returns ------- FramesMatches : New instance of the class with the selected scenes. Examples -------- .. code:: python from pprint import pprint from moviepy import * from moviepy.video.tools.cuts import FramesMatches ch_clip = VideoFileClip("media/chaplin.mp4").subclipped(1, 4) mirror_and_clip = [ch_clip.with_effects([vfx.TimeMirror()]), ch_clip] clip = concatenate_videoclips(mirror_and_clip) result = FramesMatches.from_clip(clip, 10, 3).select_scenes( 1, 2, nomatch_threshold=0, ) print(result) # [(1.0000, 4.0000, 0.0000, 0.0000), # (1.1600, 3.8400, 0.0000, 0.0000), # (1.2800, 3.7200, 0.0000, 0.0000), # (1.4000, 3.6000, 0.0000, 0.0000)] """ if nomatch_threshold is None: nomatch_threshold = match_threshold dict_starts = defaultdict(lambda: []) for start, end, min_distance, max_distance in self: dict_starts[start].append([end, min_distance, max_distance]) starts_ends = sorted(dict_starts.items(), key=lambda k: k[0]) result = [] min_start = 0 for start, ends_distances in starts_ends: if start < min_start: continue ends = [end for (end, min_distance, max_distance) in ends_distances] great_matches = [ (end, min_distance, max_distance) for (end, min_distance, max_distance) in ends_distances if max_distance < match_threshold ] great_long_matches = [ (end, min_distance, max_distance) for (end, min_distance, max_distance) in great_matches if (end - start) > min_time_span ] if not great_long_matches: continue # No GIF can be made starting at this time poor_matches = { end for (end, min_distance, max_distance) in ends_distances if min_distance > nomatch_threshold } short_matches = {end for end in ends if (end - start) <= 0.6} if not poor_matches.intersection(short_matches): continue end = max(end for (end, min_distance, max_distance) in great_long_matches) end, min_distance, max_distance = next( e for e in great_long_matches if e[0] == end ) result.append(FramesMatch(start, end, min_distance, max_distance)) min_start = start + time_distance return FramesMatches(result) def write_gifs(self, clip, gifs_dir, **kwargs): """Extract the matching frames represented by the instance from a clip and write them as GIFs in a directory, one GIF for each matching frame. Parameters ---------- clip : video.VideoClip.VideoClip A video clip whose frames scenes you want to obtain as GIF images. gif_dir : str Directory in which the GIF images will be written. kwargs Passed as ``clip.write_gif`` optional arguments. Examples -------- .. code:: python import os from pprint import pprint from moviepy import * from moviepy.video.tools.cuts import FramesMatches ch_clip = VideoFileClip("media/chaplin.mp4").subclipped(1, 4) clip = concatenate_videoclips([ch_clip.time_mirror(), ch_clip]) result = FramesMatches.from_clip(clip, 10, 3).select_scenes( 1, 2, nomatch_threshold=0, ) os.mkdir("foo") result.write_gifs(clip, "foo") # MoviePy - Building file foo/00000100_00000400.gif with imageio. # MoviePy - Building file foo/00000115_00000384.gif with imageio. # MoviePy - Building file foo/00000128_00000372.gif with imageio. # MoviePy - Building file foo/00000140_00000360.gif with imageio. """ for start, end, _, _ in self: name = "%s/%08d_%08d.gif" % (gifs_dir, 100 * start, 100 * end) clip.subclipped(start, end).write_gif(name, **kwargs) @use_clip_fps_by_default def detect_scenes( clip=None, luminosities=None, luminosity_threshold=10, logger="bar", fps=None ): """Detects scenes of a clip based on luminosity changes. Note that for large clip this may take some time. Returns ------- tuple : cuts, luminosities cuts is a series of cuts [(0,t1), (t1,t2),...(...,tf)] luminosities are the luminosities computed for each frame of the clip. Parameters ---------- clip : video.VideoClip.VideoClip, optional A video clip. Can be None if a list of luminosities is provided instead. If provided, the luminosity of each frame of the clip will be computed. If the clip has no 'fps' attribute, you must provide it. luminosities : list, optional A list of luminosities, e.g. returned by detect_scenes in a previous run. luminosity_threshold : float, optional Determines a threshold above which the 'luminosity jumps' will be considered as scene changes. A scene change is defined as a change between 2 consecutive frames that is larger than (avg * thr) where avg is the average of the absolute changes between consecutive frames. logger : str, optional Either ``"bar"`` for progress bar or ``None`` or any Proglog logger. fps : int, optional Frames per second value. Must be provided if you provide no clip or a clip without fps attribute. """ if luminosities is None: luminosities = [ f.sum() for f in clip.iter_frames(fps=fps, dtype="uint32", logger=logger) ] luminosities = np.array(luminosities, dtype=float) if clip is not None: end = clip.duration else: end = len(luminosities) * (1.0 / fps) luminosity_diffs = abs(np.diff(luminosities)) avg = luminosity_diffs.mean() luminosity_jumps = ( 1 + np.array(np.nonzero(luminosity_diffs > luminosity_threshold * avg))[0] ) timings = [0] + list((1.0 / fps) * luminosity_jumps) + [end] cuts = [(t1, t2) for t1, t2 in zip(timings, timings[1:])] return cuts, luminosities