pygame.mask

pygame module for image masks.

Useful for fast pixel perfect collision detection. A mask uses 1 bit per-pixel to store which parts collide.

pygame.mask.from_surface(Surface) -> Mask
pygame.mask.from_surface(Surface, threshold=127) -> Mask

Creates a Mask from the given surface

Creates a Mask object from the given surface by setting all the opaque pixels and not setting the transparent pixels.

If the surface uses a color-key, then it is used to decide which bits in the resulting mask are set. All the pixels that are not equal to the color-key are set and the pixels equal to the color-key are not set.

If a color-key is not used, then the alpha value of each pixel is used to decide which bits in the resulting mask are set. All the pixels that have an alpha value greater than the threshold parameter are set and the pixels with an alpha value less than or equal to the threshold are not set.

Parameters:	surface (Surface) -- the surface to create the mask from threshold (int) -- (optional) the alpha threshold (default is 127) to compare with each surface pixel's alpha value, if the surface is color-keyed this parameter is ignored
Returns:	a newly created Mask object from the given surface
Return type:	Mask

pygame.mask.from_threshold(Surface, color) -> Mask
pygame.mask.from_threshold(Surface, color, threshold=(0, 0, 0, 255), othersurface=None, palette_colors=1) -> Mask

Creates a mask by thresholding Surfaces

This is a more featureful method of getting a Mask from a surface.

If the optional othersurface is not used, all the pixels within the threshold of the color parameter are set in the resulting mask.

If the optional othersurface is used, every pixel in the first surface that is within the threshold of the corresponding pixel in othersurface is set in the resulting mask.

Parameters:	surface (Surface) -- the surface to create the mask from color (Color or int or tuple(int, int, int, [int]) or list[int, int, int, [int]]) -- color used to check if the surface's pixels are within the given threshold range, this parameter is ignored if the optional othersurface parameter is supplied threshold (Color or int or tuple(int, int, int, [int]) or list[int, int, int, [int]]) -- (optional) the threshold range used to check the difference between two colors (default is (0, 0, 0, 255)) othersurface (Surface) -- (optional) used to check whether the pixels of the first surface are within the given threshold range of the pixels from this surface (default is None) palette_colors (int) -- (optional) indicates whether to use the palette colors or not, a nonzero value causes the palette colors to be used and a 0 causes them not to be used (default is 1)
Returns:	a newly created Mask object from the given surface
Return type:	Mask

pygame.mask.Mask

pygame object for representing 2D bitmasks

Mask(size=(width, height)) -> Mask

Mask(size=(width, height), fill=False) -> Mask

A Mask object is used to represent a 2D bitmask. Each bit in the mask represents a pixel. 1 is used to indicate a set bit and 0 is used to indicate an unset bit. Set bits in a mask can be used to detect collisions with other masks and their set bits.

A filled mask has all of its bits set to 1, conversely an unfilled/cleared/empty mask has all of its bits set to 0. Masks can be created unfilled (default) or filled by using the fill parameter. Masks can also be cleared or filled using the pygame.mask.Mask.clear() and pygame.mask.Mask.fill() methods respectively.

A mask's coordinates start in the top left corner at (0, 0) just like pygame.Surface. Individual bits can be accessed using the pygame.mask.Mask.get_at() and pygame.mask.Mask.set_at() methods.

The methods overlap(), overlap_area(), overlap_mask(), draw(), erase(), and convolve() use an offset parameter to indicate the offset of another mask's top left corner from the calling mask's top left corner. The calling mask's top left corner is considered to be the origin (0, 0). Offsets are a tuple or list of 2 integer values (x_offset, y_offset). Positive and negative offset values are supported.

           0 to x (x_offset)
           :    :
   0 ..... +----:---------+
   to      |    :         |
   y .......... +-----------+
(y_offset) |    | othermask |
           |    +-----------+
           | calling_mask |
           +--------------+

Parameters:	size (tuple(int, int) or list[int, int]) -- the dimensions of the mask (width and height) fill (bool) -- (optional) create an unfilled mask (default: False) or filled mask (True)
Returns:	a newly created Mask object
Return type:	Mask

Changed in pygame 2.0.0: Shallow copy support added. The Mask class supports the special method __copy__() and shallow copying via copy.copy(mask).

Changed in pygame 2.0.0: Subclassing support added. The Mask class can be used as a base class.

Changed in pygame 1.9.5: Added support for keyword arguments.

Changed in pygame 1.9.5: Added the optional keyword parameter fill.

Changed in pygame 1.9.5: Added support for masks with a width and/or a height of 0.

copy() -> Mask

Returns a new copy of the mask

Returns:	a new copy of this mask, the new mask will have the same width, height, and set/unset bits as the original
Return type:	Mask

Note

If a mask subclass needs to copy any instance specific attributes then it should override the __copy__() method. The overridden __copy__() method needs to call super().__copy__() and then copy the required data as in the following example code.

class SubMask(pygame.mask.Mask):
    def __copy__(self):
        new_mask = super().__copy__()
        # Do any SubMask attribute copying here.
        return new_mask

New in pygame 2.0.0.

get_size() -> (width, height)

Returns the size of the mask

Returns:	the size of the mask, (width, height)
Return type:	tuple(int, int)

get_rect(**kwargs) -> Rect

Returns a Rect based on the size of the mask

Returns a new pygame.Rect() object based on the size of this mask. The rect's default position will be (0, 0) and its default width and height will be the same as this mask's. The rect's attributes can be altered via pygame.Rect() attribute keyword arguments/values passed into this method. As an example, a_mask.get_rect(center=(10, 5)) would create a pygame.Rect() based on the mask's size centered at the given position.

Parameters:	kwargs (dict) -- pygame.Rect() attribute keyword arguments/values that will be applied to the rect
Returns:	a new pygame.Rect() object based on the size of this mask with any pygame.Rect() attribute keyword arguments/values applied to it
Return type:	Rect

New in pygame 2.0.0.

get_at((x, y)) -> int

Gets the bit at the given position

Parameters:	pos (tuple(int, int) or list[int, int]) -- the position of the bit to get
Returns:	1 if the bit is set, 0 if the bit is not set
Return type:	int
Raises:	IndexError -- if the position is outside of the mask's bounds

set_at((x, y)) -> None
set_at((x, y), value=1) -> None

Sets the bit at the given position

Parameters:	pos (tuple(int, int) or list[int, int]) -- the position of the bit to set value (int) -- any nonzero int will set the bit to 1, 0 will set the bit to 0 (default is 1)
Returns:	None
Return type:	NoneType
Raises:	IndexError -- if the position is outside of the mask's bounds

overlap(othermask, offset) -> (x, y)
overlap(othermask, offset) -> None

Returns the point of intersection

Returns the first point of intersection encountered between this mask and othermask. A point of intersection is 2 overlapping set bits.

The current algorithm searches the overlapping area in sizeof(unsigned long int) * CHAR_BIT bit wide column blocks (the value of sizeof(unsigned long int) * CHAR_BIT is platform dependent, for clarity it will be referred to as W). Starting at the top left corner it checks bits 0 to W - 1 of the first row ((0, 0) to (W - 1, 0)) then continues to the next row ((0, 1) to (W - 1, 1)). Once this entire column block is checked, it continues to the next one (W to 2 * W - 1). This is repeated until it finds a point of intersection or the entire overlapping area is checked.

Parameters:	othermask (Mask) -- the other mask to overlap with this mask offset (tuple(int, int) or list[int, int]) -- the offset of othermask from this mask, for more details refer to the Mask offset notes
Returns:	point of intersection or None if no intersection
Return type:	tuple(int, int) or NoneType

overlap_area(othermask, offset) -> numbits

Returns the number of overlapping set bits

Returns the number of overlapping set bits between between this mask and othermask.

This can be useful for collision detection. An approximate collision normal can be found by calculating the gradient of the overlapping area through the finite difference.

dx = mask.overlap_area(othermask, (x + 1, y)) - mask.overlap_area(othermask, (x - 1, y))
dy = mask.overlap_area(othermask, (x, y + 1)) - mask.overlap_area(othermask, (x, y - 1))

Parameters:	othermask (Mask) -- the other mask to overlap with this mask offset (tuple(int, int) or list[int, int]) -- the offset of othermask from this mask, for more details refer to the Mask offset notes
Returns:	the number of overlapping set bits
Return type:	int

overlap_mask(othermask, offset) -> Mask

Returns a mask of the overlapping set bits

Returns a Mask, the same size as this mask, containing the overlapping set bits between this mask and othermask.

Parameters:	othermask (Mask) -- the other mask to overlap with this mask offset (tuple(int, int) or list[int, int]) -- the offset of othermask from this mask, for more details refer to the Mask offset notes
Returns:	a newly created Mask with the overlapping bits set
Return type:	Mask

fill() -> None

Sets all bits to 1

Sets all bits in the mask to 1.

Returns:	None
Return type:	NoneType

clear() -> None

Sets all bits to 0

Sets all bits in the mask to 0.

Returns:	None
Return type:	NoneType

invert() -> None

Flips all the bits

Flips all of the bits in the mask. All the set bits are cleared to 0 and all the unset bits are set to 1.

Returns:	None
Return type:	NoneType

scale((width, height)) -> Mask

Resizes a mask

Creates a new Mask of the requested size with its bits scaled from this mask.

Parameters:	size (tuple(int, int) or list[int, int]) -- the width and height (size) of the mask to create
Returns:	a new Mask object with its bits scaled from this mask
Return type:	Mask
Raises:	ValueError -- if width < 0 or height < 0

draw(othermask, offset) -> None

Draws a mask onto another

Performs a bitwise OR, drawing othermask onto this mask.

Parameters:	othermask (Mask) -- the mask to draw onto this mask offset (tuple(int, int) or list[int, int]) -- the offset of othermask from this mask, for more details refer to the Mask offset notes
Returns:	None
Return type:	NoneType

erase(othermask, offset) -> None

Erases a mask from another

Erases (clears) all bits set in othermask from this mask.

Parameters:	othermask (Mask) -- the mask to erase from this mask offset (tuple(int, int) or list[int, int]) -- the offset of othermask from this mask, for more details refer to the Mask offset notes
Returns:	None
Return type:	NoneType

count() -> bits

Returns the number of set bits

Returns:	the number of set bits in the mask
Return type:	int

centroid() -> (x, y)

Returns the centroid of the set bits

Finds the centroid (the center mass of the set bits) for this mask.

Returns:	a coordinate tuple indicating the centroid of the mask, it will return (0, 0) if the mask has no bits set
Return type:	tuple(int, int)

angle() -> theta

Returns the orientation of the set bits

Finds the approximate orientation (from -90 to 90 degrees) of the set bits in the mask. This works best if performed on a mask with only one connected component.

Returns:	the orientation of the set bits in the mask, it will return 0.0 if the mask has no bits set
Return type:	float

Note

See connected_component() for details on how a connected component is calculated.

outline() -> [(x, y), ...]
outline(every=1) -> [(x, y), ...]

Returns a list of points outlining an object

Returns a list of points of the outline of the first connected component encountered in the mask. To find a connected component, the mask is searched per row (left to right) starting in the top left corner.

The every optional parameter skips set bits in the outline. For example, setting it to 10 would return a list of every 10th set bit in the outline.

Parameters:	every (int) -- (optional) indicates the number of bits to skip over in the outline (default is 1)
Returns:	a list of points outlining the first connected component encountered, an empty list is returned if the mask has no bits set
Return type:	list[tuple(int, int)]

Note

See connected_component() for details on how a connected component is calculated.

convolve(othermask) -> Mask
convolve(othermask, outputmask=None, offset=(0, 0)) -> Mask

Returns the convolution of this mask with another mask

Convolve this mask with the given othermask.

Parameters:	othermask (Mask) -- mask to convolve this mask with outputmask (Mask or NoneType) -- (optional) mask for output (default is None) offset (tuple(int, int) or list[int, int]) -- the offset of othermask from this mask, (default is (0, 0))
Returns:	a Mask with the (i - offset[0], j - offset[1]) bit set, if shifting othermask (such that its bottom right corner is at (i, j)) causes it to overlap with this mask If an outputmask is specified, the output is drawn onto it and it is returned. Otherwise a mask of size (MAX(0, width + othermask's width - 1), MAX(0, height + othermask's height - 1)) is created and returned.
Return type:	Mask

connected_component() -> Mask
connected_component((x, y)) -> Mask

Returns a mask containing a connected component

A connected component is a group (1 or more) of connected set bits (orthogonally and diagonally). The SAUF algorithm, which checks 8 point connectivity, is used to find a connected component in the mask.

By default this method will return a Mask containing the largest connected component in the mask. Optionally, a bit coordinate can be specified and the connected component containing it will be returned. If the bit at the given location is not set, the returned Mask will be empty (no bits set).

Parameters:	pos (tuple(int, int) or list[int, int]) -- (optional) selects the connected component that contains the bit at this position
Returns:	a Mask object (same size as this mask) with the largest connected component from this mask, if this mask has no bits set then an empty mask will be returned If the pos parameter is provided then the mask returned will have the connected component that contains this position. An empty mask will be returned if the pos parameter selects an unset bit.
Return type:	Mask
Raises:	IndexError -- if the optional pos parameter is outside of the mask's bounds

connected_components() -> [Mask, ...]
connected_components(min=0) -> [Mask, ...]

Returns a list of masks of connected components

Provides a list containing a Mask object for each connected component.

Parameters:	min (int) -- (optional) indicates the minimum number of bits (to filter out noise) per connected component (default is 0, which equates to no minimum and is equivalent to setting it to 1, as a connected component must have at least 1 bit set)
Returns:	a list containing a Mask object for each connected component, an empty list is returned if the mask has no bits set
Return type:	list[Mask]

Note

See connected_component() for details on how a connected component is calculated.

get_bounding_rects() -> [Rect, ...]

Returns a list of bounding rects of connected components

Provides a list containing a bounding rect for each connected component.

Returns:	a list containing a bounding rect for each connected component, an empty list is returned if the mask has no bits set
Return type:	list[Rect]

Note

See connected_component() for details on how a connected component is calculated.

to_surface() -> Surface
to_surface(surface=None, setsurface=None, unsetsurface=None, setcolor=(255, 255, 255, 255), unsetcolor=(0, 0, 0, 255)) -> Surface

Returns a surface with the mask drawn on it

Draws this mask on the given surface. Set bits (bits set to 1) and unset bits (bits set to 0) can be drawn onto a surface.

Parameters:	surface (Surface or None) -- (optional) Surface to draw mask onto, if no surface is provided one will be created (default is None, which will cause a surface with the parameters Surface(size=mask.get_size(), flags=SRCALPHA, depth=32) to be created, drawn on, and returned) setsurface (Surface or None) -- (optional) use this surface's color values to draw set bits (default is None), if this surface is smaller than the mask any bits outside its bounds will use the setcolor value unsetsurface (Surface or None) -- (optional) use this surface's color values to draw unset bits (default is None), if this surface is smaller than the mask any bits outside its bounds will use the unsetcolor value setcolor (Color or int or tuple(int, int, int, [int]) or list(int, int, int, [int]) or None) -- (optional) color to draw set bits (default is (255, 255, 255, 255), white), use None to skip drawing the set bits, the setsurface parameter (if set) will takes precedence over this parameter unsetcolor (Color or int or tuple(int, int, int, [int]) or list(int, int, int, [int]) or None) -- (optional) color to draw unset bits (default is (0, 0, 0, 255), black), use None to skip drawing the unset bits, the unsetsurface parameter (if set) will takes precedence over this parameter
Returns:	the surface parameter (or a newly created surface if no surface parameter was provided) with this mask drawn on it
Return type:	Surface

Note

To skip drawing the set bits, both setsurface and setcolor must be None. The setsurface parameter defaults to None, but setcolor defaults to a color value and therefore must be set to None.

Note

To skip drawing the unset bits, both unsetsurface and unsetcolor must be None. The unsetsurface parameter defaults to None, but unsetcolor defaults to a color value and therefore must be set to None.

New in pygame 2.0.0.

---

Edit on GitHub