1. algorithms
  2. tpereira
  3. lbp_extractor
  4. 1

LBP Feature extractor

This algorithm is a legacy one. The API has changed since its implementation. New versions and forks will need to be updated.

Algorithms have at least one input and one output. All algorithm endpoints are organized in groups. Groups are used by the platform to indicate which inputs and outputs are synchronized together. The first group is automatically synchronized with the channel defined by the block in which the algorithm is deployed.

Unnamed group

Endpoint Name Data Format Nature
image system/array_3d_uint8/1 Input
eye_centers system/eye_positions/1 Input
histogram system/array_1d_floats/1 Output

Parameters allow users to change the configuration of an algorithm when scheduling an experiment

Name Description Type Default Range/Choices
block_size int32 12
block_overlap int32 0
cropped_image_height uint32 80
cropped_image_width uint32 64
right_eye_pos_x uint32 16
right_eye_pos_y uint32 15
left_eye_pos_x uint32 16
left_eye_pos_y uint32 48
lbp_P int32 8
lbp_R float64 2.0 
xxxxxxxxxx
80
 
1
import bob
2
import numpy
3
​
4
​
5
​
6
"""
7
Receive an image and computes the LBP histogram
8
​
9
LBP always squared
10
​
11
"""
12
​
13
class Algorithm:
14
​
15
  def __init__(self):
16
    self.block_size           = 12
17
    self.block_overlap        = 0
18
    self.cropped_image_height = 80
19
    self.cropped_image_width  = 64
20
    self.right_eye_pos_x      = 16
21
    self.right_eye_pos_y      = 15
22
    self.left_eye_pos_x       = 16
23
    self.left_eye_pos_y       = 48
24
    self.lbp_P                = 8
25
    self.lbp_R                = 1.
26
​
27
​
28
  def setup(self, parameters):
29
    self.block_size           = parameters.get('block_size', self.block_size)
30
    self.block_overlap        = parameters.get('block_overlap', self.block_overlap)
31
    self.cropped_image_height = parameters.get('cropped_image_height', self.cropped_image_height)
32
    self.cropped_image_width  = parameters.get('cropped_image_width', self.cropped_image_width)
33
    self.right_eye_pos_x      = parameters.get('right_eye_pos_x', self.right_eye_pos_x)
34
    self.right_eye_pos_y      = parameters.get('right_eye_pos_y', self.right_eye_pos_y)
35
    self.left_eye_pos_x       = parameters.get('left_eye_pos_x', self.left_eye_pos_x)
36
    self.left_eye_pos_y       = parameters.get('left_eye_pos_y', self.left_eye_pos_y)
37
    self.lbp_P                = parameters.get('lbp_P', self.lbp_P)
38
    self.lbp_R                = parameters.get('lbp_R', self.lbp_R)
39
​
40
    self.face_norm = bob.ip.FaceEyesNorm(
41
          self.cropped_image_height, # cropped image height
42
          self.cropped_image_width, # cropped image width
43
          self.right_eye_pos_y, # Y of first position (usually: right eye)
44
          self.right_eye_pos_x, # X of first position (usually: right eye)
45
          self.left_eye_pos_y,  # Y of second position (usually: left eye)
46
          self.left_eye_pos_x   # X of second position (usually: left eye)
47
    )
48
  
49
    self.lbp = bob.ip.LBPHSFeatures(self.block_size,
50
                                    self.block_size,
51
                                    self.block_overlap,
52
                                    self.block_overlap,
53
                                    bob.ip.LBP(self.lbp_P, self.lbp_R, uniform = True)
54
                                   )
55
​
56
    return True
57
​
58
  def process(self, inputs, outputs):
59
    image = inputs["image"].data.value
60
    image = bob.ip.rgb_to_gray(image)
61
​
62
    eye_positions = inputs["eye_centers"].data
63
​
64
    annotation_right_eye_y = int(eye_positions.right.y)
65
    annotation_right_eye_x = int(eye_positions.right.x)
66
​
67
    annotation_left_eye_y = int(eye_positions.left.y)
68
    annotation_left_eye_x = int(eye_positions.left.x)
69
​
70
    normalize_face = self.face_norm(image, annotation_right_eye_y, annotation_right_eye_x, annotation_left_eye_y, annotation_left_eye_x)
71
    histogram      = numpy.hstack(self.lbp(normalize_face)).astype(numpy.float64)
72
    #import ipdb; ipdb.set_trace();
73
​
74
    outputs["histogram"].write({
75
    'value':histogram
76
    })
77
​
78
​
79
    return True
80
​

The code for this algorithm in Python
The ruler at 80 columns indicate suggested POSIX line breaks (for readability).
The editor will automatically enlarge to accomodate the entirety of your input
Use keyboard shortcuts for search/replace and faster editing. For example, use Ctrl-F (PC) or Cmd-F (Mac) to search through this box

This algorithm computes:
  • A face crop, given the eyes center coordinates.
  • LBP histogram for each block (patch). The histograms for each block are then concatenated.

Details about LBP operator can be found in [Ojala02]

The inputs are:
  • image: RGB image
  • eye_centers: The coordinates of the eyes.

The output are the concatenated histograms of each block.

[Ojala02]
  1. Ojala, M. Pietikainen, T. Maenpaa: Multiresolution gray-scale and rotation invariant texture classification with Local Binary Patterns. IEEE Transactions on Pattern Analysis and Machine inteligence 2002
No experiments are using this algorithm.
Created with Raphaël 2.1.2[compare]tpereira/lbp_extractor/12014Aug28

This table shows the number of times this algorithm has been successfully run using the given environment. Note this does not provide sufficient information to evaluate if the algorithm will run when submitted to different conditions.

Terms of Service | Contact Information | BEAT platform version 2.2.1b0 | © Idiap Research Institute - 2013-2025