Skip to content

Masks Utils

move_masks

supervision.detection.utils.masks.move_masks(masks, offset, resolution_wh)

Offset the masks in an array by the specified (x, y) amount.

Parameters:

Name Type Description Default

masks

NDArray[bool_]

A 3D array of binary masks corresponding to the predictions. Shape: (N, H, W), where N is the number of predictions, and H, W are the dimensions of each mask.

 required

offset

NDArray[int32]

An array of shape (2,) containing int values [dx, dy]. Supports both positive and negative values for bidirectional movement.

 required

resolution_wh

Tuple[int, int]

The width and height of the desired mask resolution.

 required

Returns:

Type Description
NDArray[bool_]

(npt.NDArray[np.bool_]) repositioned masks, optionally padded to the specified shape.

Examples:

import numpy as np
import supervision as sv

mask = np.array([[[False, False, False, False],
                 [False, True,  True,  False],
                 [False, True,  True,  False],
                 [False, False, False, False]]], dtype=bool)

offset = np.array([1, 1])
sv.move_masks(mask, offset, resolution_wh=(4, 4))
# array([[[False, False, False, False],
#         [False, False, False, False],
#         [False, False,  True,  True],
#         [False, False,  True,  True]]], dtype=bool)

offset = np.array([-2, 2])
sv.move_masks(mask, offset, resolution_wh=(4, 4))
# array([[[False, False, False, False],
#         [False, False, False, False],
#         [False, False, False, False],
#         [True,  False, False, False]]], dtype=bool)
Source code in supervision/detection/utils/masks.py 
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
def move_masks(
    masks: npt.NDArray[np.bool_],
    offset: npt.NDArray[np.int32],
    resolution_wh: tuple[int, int],
) -> npt.NDArray[np.bool_]:
    """
    Offset the masks in an array by the specified (x, y) amount.

    Args:
        masks (npt.NDArray[np.bool_]): A 3D array of binary masks corresponding to the
            predictions. Shape: `(N, H, W)`, where N is the number of predictions, and
            H, W are the dimensions of each mask.
        offset (npt.NDArray[np.int32]): An array of shape `(2,)` containing int values
            `[dx, dy]`. Supports both positive and negative values for bidirectional
            movement.
        resolution_wh (Tuple[int, int]): The width and height of the desired mask
            resolution.

    Returns:
        (npt.NDArray[np.bool_]) repositioned masks, optionally padded to the specified
            shape.

    Examples:
        ```python
        import numpy as np
        import supervision as sv

        mask = np.array([[[False, False, False, False],
                         [False, True,  True,  False],
                         [False, True,  True,  False],
                         [False, False, False, False]]], dtype=bool)

        offset = np.array([1, 1])
        sv.move_masks(mask, offset, resolution_wh=(4, 4))
        # array([[[False, False, False, False],
        #         [False, False, False, False],
        #         [False, False,  True,  True],
        #         [False, False,  True,  True]]], dtype=bool)

        offset = np.array([-2, 2])
        sv.move_masks(mask, offset, resolution_wh=(4, 4))
        # array([[[False, False, False, False],
        #         [False, False, False, False],
        #         [False, False, False, False],
        #         [True,  False, False, False]]], dtype=bool)
        ```
    """
    mask_array = np.full((masks.shape[0], resolution_wh[1], resolution_wh[0]), False)

    if offset[0] < 0:
        source_x_start = -offset[0]
        source_x_end = min(masks.shape[2], resolution_wh[0] - offset[0])
        destination_x_start = 0
        destination_x_end = min(resolution_wh[0], masks.shape[2] + offset[0])
    else:
        source_x_start = 0
        source_x_end = min(masks.shape[2], resolution_wh[0] - offset[0])
        destination_x_start = offset[0]
        destination_x_end = offset[0] + source_x_end - source_x_start

    if offset[1] < 0:
        source_y_start = -offset[1]
        source_y_end = min(masks.shape[1], resolution_wh[1] - offset[1])
        destination_y_start = 0
        destination_y_end = min(resolution_wh[1], masks.shape[1] + offset[1])
    else:
        source_y_start = 0
        source_y_end = min(masks.shape[1], resolution_wh[1] - offset[1])
        destination_y_start = offset[1]
        destination_y_end = offset[1] + source_y_end - source_y_start

    if source_x_end > source_x_start and source_y_end > source_y_start:
        mask_array[
            :,
            destination_y_start:destination_y_end,
            destination_x_start:destination_x_end,
        ] = masks[:, source_y_start:source_y_end, source_x_start:source_x_end]

    return mask_array

contains_holes

supervision.detection.utils.masks.contains_holes(mask)

Checks if the binary mask contains holes (background pixels fully enclosed by foreground pixels).

Parameters:

Name Type Description Default

mask

NDArray[bool_]

2D binary mask where True indicates foreground object and False indicates background.

 required

Returns:

Type Description
bool

True if holes are detected, False otherwise.

Examples:

import numpy as np
import supervision as sv

mask = np.array([
    [0, 0, 0, 0, 0],
    [0, 1, 1, 1, 0],
    [0, 1, 0, 1, 0],
    [0, 1, 1, 1, 0],
    [0, 0, 0, 0, 0]
]).astype(bool)

sv.contains_holes(mask=mask)
# True

mask = np.array([
    [0, 0, 0, 0, 0],
    [0, 1, 1, 1, 0],
    [0, 1, 1, 1, 0],
    [0, 1, 1, 1, 0],
    [0, 0, 0, 0, 0]
]).astype(bool)

sv.contains_holes(mask=mask)
# False

contains_holes

Source code in supervision/detection/utils/masks.py 
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
def contains_holes(mask: npt.NDArray[np.bool_]) -> bool:
    """
    Checks if the binary mask contains holes (background pixels fully enclosed by
    foreground pixels).

    Args:
        mask (npt.NDArray[np.bool_]): 2D binary mask where `True` indicates foreground
            object and `False` indicates background.

    Returns:
        True if holes are detected, False otherwise.

    Examples:
        ```python
        import numpy as np
        import supervision as sv

        mask = np.array([
            [0, 0, 0, 0, 0],
            [0, 1, 1, 1, 0],
            [0, 1, 0, 1, 0],
            [0, 1, 1, 1, 0],
            [0, 0, 0, 0, 0]
        ]).astype(bool)

        sv.contains_holes(mask=mask)
        # True

        mask = np.array([
            [0, 0, 0, 0, 0],
            [0, 1, 1, 1, 0],
            [0, 1, 1, 1, 0],
            [0, 1, 1, 1, 0],
            [0, 0, 0, 0, 0]
        ]).astype(bool)

        sv.contains_holes(mask=mask)
        # False
        ```

    ![contains_holes](https://media.roboflow.com/supervision-docs/contains-holes.png){ align=center width="800" }
    """  # noqa E501 // docs
    mask_uint8 = mask.astype(np.uint8)
    _, hierarchy = cv2.findContours(mask_uint8, cv2.RETR_CCOMP, cv2.CHAIN_APPROX_SIMPLE)

    if hierarchy is not None:
        parent_contour_index = 3
        for h in hierarchy[0]:
            if h[parent_contour_index] != -1:
                return True
    return False

contains_multiple_segments

supervision.detection.utils.masks.contains_multiple_segments(mask, connectivity=4)

Checks if the binary mask contains multiple unconnected foreground segments.

Parameters:

Name Type Description Default

mask

NDArray[bool_]

2D binary mask where True indicates foreground object and False indicates background.

 required

connectivity

int)

Default: 4 is 4-way connectivity, which means that foreground pixels are the part of the same segment/component if their edges touch. Alternatively: 8 for 8-way connectivity, when foreground pixels are connected by their edges or corners touch.

 4

Returns:

Type Description
bool

True when the mask contains multiple not connected components, False otherwise.

Raises:

Type Description
ValueError

If connectivity(int) parameter value is not 4 or 8.

Examples:

import numpy as np
import supervision as sv

mask = np.array([
    [0, 0, 0, 0, 0, 0],
    [0, 1, 1, 0, 1, 1],
    [0, 1, 1, 0, 1, 1],
    [0, 0, 0, 0, 0, 0],
    [0, 1, 1, 1, 0, 0],
    [0, 1, 1, 1, 0, 0]
]).astype(bool)

sv.contains_multiple_segments(mask=mask, connectivity=4)
# True

mask = np.array([
    [0, 0, 0, 0, 0, 0],
    [0, 1, 1, 1, 1, 1],
    [0, 1, 1, 1, 1, 1],
    [0, 1, 1, 1, 1, 1],
    [0, 1, 1, 1, 1, 1],
    [0, 0, 0, 0, 0, 0]
]).astype(bool)

sv.contains_multiple_segments(mask=mask, connectivity=4)
# False

contains_multiple_segments

Source code in supervision/detection/utils/masks.py 
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
def contains_multiple_segments(
    mask: npt.NDArray[np.bool_], connectivity: int = 4
) -> bool:
    """
    Checks if the binary mask contains multiple unconnected foreground segments.

    Args:
        mask (npt.NDArray[np.bool_]): 2D binary mask where `True` indicates foreground
            object and `False` indicates background.
        connectivity (int) : Default: 4 is 4-way connectivity, which means that
            foreground pixels are the part of the same segment/component
            if their edges touch.
            Alternatively: 8 for 8-way connectivity, when foreground pixels are
            connected by their edges or corners touch.

    Returns:
        True when the mask contains multiple not connected components, False otherwise.

    Raises:
        ValueError: If connectivity(int) parameter value is not 4 or 8.

    Examples:
        ```python
        import numpy as np
        import supervision as sv

        mask = np.array([
            [0, 0, 0, 0, 0, 0],
            [0, 1, 1, 0, 1, 1],
            [0, 1, 1, 0, 1, 1],
            [0, 0, 0, 0, 0, 0],
            [0, 1, 1, 1, 0, 0],
            [0, 1, 1, 1, 0, 0]
        ]).astype(bool)

        sv.contains_multiple_segments(mask=mask, connectivity=4)
        # True

        mask = np.array([
            [0, 0, 0, 0, 0, 0],
            [0, 1, 1, 1, 1, 1],
            [0, 1, 1, 1, 1, 1],
            [0, 1, 1, 1, 1, 1],
            [0, 1, 1, 1, 1, 1],
            [0, 0, 0, 0, 0, 0]
        ]).astype(bool)

        sv.contains_multiple_segments(mask=mask, connectivity=4)
        # False
        ```

    ![contains_multiple_segments](https://media.roboflow.com/supervision-docs/contains-multiple-segments.png){ align=center width="800" }
    """  # noqa E501 // docs
    if connectivity != 4 and connectivity != 8:
        raise ValueError(
            "Incorrect connectivity value. Possible connectivity values: 4 or 8."
        )
    mask_uint8 = mask.astype(np.uint8)
    labels = np.zeros_like(mask_uint8, dtype=np.int32)
    number_of_labels, _ = cv2.connectedComponents(
        mask_uint8, labels, connectivity=connectivity
    )
    return number_of_labels > 2

Comments