Coverage for brainbox/video.py: 96%

1"""Functions for analyzing video frame data""" 

2import numpy as np (empty)

3import cv2 (empty)

4 

5 

6def frame_diff(frame1, frame2): (empty)

7 """ 

8 Outputs pythagorean distance between two frames 

9 :param frame1: A numpy array of pixels with a shape of either (m, n, 3) or (m, n) 

10 :param frame2: A numpy array of pixels with a shape of either (m, n, 3) or (m, n) 

11 :return: An array with a shape equal to the input frames 

12 """ 

13 if frame1.shape != frame2.shape: 1 ctx1e

14 raise ValueError('Frames must have the same shape') 1 ctx1e

15 diff32 = np.float32(frame1) - np.float32(frame2) 1 ctx1e

16 if frame1.ndim == 3: 1 ctx1e

17 norm32 = (np.sqrt(diff32[:, :, 0] ** 2 + diff32[:, :, 1] ** 2 + diff32[:, :, 2] ** 2) / 1 ctx1e

18 np.sqrt(255 ** 2 * 3)) 

19 else: 

20 norm32 = np.sqrt(diff32 ** 2 * 3) / np.sqrt(255 ** 2 * 3) 1 ctx1e

21 return np.uint8(norm32 * 255) 1 ctx1e

22 

23 

24def frame_diffs(frames, diff=1): (empty)

25 """ 

26 Return the difference between frames. May also take difference between more than 1 frames. 

27 Values are normalized between 0-255. 

28 :param frames: Array or list of frames, where each frame is either (y, x) or (y, x, 3). 

29 :param diff: Take difference between frames N and frames N + diff. 

30 :return: uint8 array with shape (n-diff, y, x). 

31 """ 

32 frames = np.array(frames, dtype=np.float32) 4 ctx1dabc

33 if frames.shape[0] < diff: 4 ctx1dabc

34 raise ValueError('Difference must be less than number of frames') 1 ctx1d

35 diff32 = frames[diff:] - frames[:-diff] 4 ctx1dabc

36 # Normalize 

37 if frames.ndim == 4: 4 ctx1dabc

38 norm32 = np.sqrt((diff32 ** 2).sum(axis=3)) / np.sqrt(255 ** 2 * 3) 1 ctx1d

39 else: 

40 norm32 = np.sqrt(diff32 ** 2 * 3) / np.sqrt(255 ** 2 * 3) 4 ctx1dabc

41 return np.uint8(norm32 * 255) 4 ctx1dabc

42 

43 

44def motion_energy(frames, diff=2, kernel=None, normalize=True): (empty)

45 """ 

46 Returns a min-max normalized vector of motion energy between frames. 

47 :param frames: A list of ndarray of frames. 

48 :param diff: Take difference between frames N and frames N + diff. 

49 :param kernel: An optional Gaussian smoothing to apply with a given kernel size. 

50 :param normalize: If True, motion energy is min-max normalized 

51 :return df_: A vector of length n frames - diff, normalized between 0 and 1. 

52 :return stDev: The standard deviation between the frames (not normalized). 

53 

54 Example 1 - Calculate normalized difference between consecutive frames 

55 df, std = motion_energy(frames, diff=1) 

56 

57 Example 2 - Calculate smoothed difference between every 2nd frame 

58 df, _ = motion_energy(frames, kernel=(9, 9)) 

59 """ 

60 df = frame_diffs(frames, diff) 3 ctx1abc

61 

62 # Smooth with a Gaussian blur TODO Use median blur instead 

63 if kernel is not None: 3 ctx1abc

64 df = cv2.GaussianBlur(df, (9, 9), 0) 

65 stDev = np.array([cv2.meanStdDev(x)[1] for x in df]).squeeze() 3 ctx1abc

66 

67 # Feature scaling 

68 df_ = df.sum(axis=(1, 2)) 3 ctx1abc

69 if normalize: 3 ctx1abc

70 df_ = (df_ - df_.min()) / (df_.max() - df_.min()) 3 ctx1abc

71 return df_, stDev 3 ctx1abc