sroc 0.2.0

Straight Rectangles on Canvas

SROC

Straight Rectangles on Canvas

SROC (Straight Rectangles on Canvas) is a Python library for manipulating straight rectangles and handling interactions with a canvas. It provides classes and methods for creating, modifying, and analyzing rectangles, as well as handling image tiling and canvas scanning operations.

Features

• Rect: A class to represent a rectangle with various utility methods for manipulation and calculation.
• Rects: Base class to manage a collection of Rect objects with various utility methods for manipulation and calculation.
• CactusRects: Specialized class derived from Rects to create and handle collections of adjacent rectangles.
• RectCanvas: Base class to represent a canvas with overlayed rectangles and provide various utility methods for manipulation and visualization.
• RectScanner: Specialized class derived from RectCanvas to scan and detect filled rectangles from the underlying canvas and extract as rectangles.
• CanvasTiler: Specialized class derived from RectCanvas to handle tiling operations on the canvas.

Usage

Rect Class

The Rect class represents a rectangle and provides various methods for manipulation and calculation.

from sroc import Rect

# Create a rectangle from a bounding box (x0, y0, x1, y1)
# Same as Rect().fromBox()
rect1 = Rect(box=(0, 0, 100, 100))
# Create a rectangle from a bounding box with specific order
# ('xy' is the default, 'yx, 'xx', 'yy' are supported)
rect1 = Rect(box=(0, 0, 100, 100), order="yx")

# Create a rectangle from tuple of n points (y1, x1, y0, x0)
# Same as Rect().fromPoints(). 
rect2 = Rect(points=((0, 0), (100, 100)))

# Get the width and height of the rectangle
width = rect1.width()
height = rect1.height()

# Rotate the rectangle by 90 degrees
rect1.rotate(90)

# Find distance between rect1 and rect2
distance = rect1.getDistFromRect(rect2, reference='border', type="manhattan")

# Get a tuple with the 4 vertex as y,x points
rect1_points = rect1.to4points(order="yx", sequence="maxmin")

Viewport

Rectangle viewport can be set both at initialization passing vp_xmax and vp_ymax or calling setViewportLimit(vp_xmax, vp_ymax). Rectangle coordinates can be set or retrieved as fractions of the viewport setting use_vp_fractions=True.

Rects Class

The Rects class manages a collection of Rect objects and provides methods for manipulation and calculation.

from sroc import Rect
from sroc import Rects

# Create a collection of rectangles
rects1 = Rects()

# Add a rectangle to the collection
rects1.addRect(Rect(box=(0, 0, 100, 100)))

# Add a rectangle from bounding box or tuples of n points (same methods from Rect can be used)
rects2 = Rects().fromBox([(0, 0, 100, 100), (0, 100, 100, 200)])

# Simplify the collection of rectangles (will merge rects to (0, 0, 100, 200))
rects2.simplifyRects(tolerance=5)

# Moves rectangles from rect2 to rect1
rects1.moveRectsFrom(rects2)

# Get a list of the rectangles as 'xy' 'minmax' bounding boxes
rectangles = rects1.toBox()

CactusRects Class

The CactusRects class is a specialized class derived from Rects to handle specific operations with tolerance and strategy.

from sroc import CactusRects

# Create a CactusRects object with a seed rectangle
cactus_rects = CactusRects(seedRect=Rect(box=(0, 0, 100, 100)), tolerance=5, strategy="full")

# Add a rectangle to the collection
cactus_rects.addRect(Rect(box=(50, 50, 150, 150)))

RectCanvas Class

The RectCanvas class represents a canvas with rectangles and provides various methods for manipulation and visualization.

from sroc import RectCanvas

# Create a RectCanvas object from an image path
canvas = RectCanvas(image_path="path/to/image.png")

# Get the original canvas
original_canvas = canvas.getOriginalCanvas()

# Get the current canvas with modifications
current_canvas = canvas.getCurrentCanvas(show_rects=True, show_labels=True)

RectScanner Class

The RectScanner class is derived from RectCanvas to scan and detect filled rectangles in an image.

from sroc.overlay import RectScanner

# Create a RectScanner object from an image path
scanner = RectScanner(image_path="path/to/image.png")

# Scan the image for filled rectangles
filled_rects = scanner.scanFilledRects(color_treshold=200, min_lenght=20)

CanvasTiler Class

The CanvasTiler class is derived from RectCanvas to handle tiling operations on the canvas.

from sroc.overlay import CanvasTiler

# Create a CanvasTiler object from an image path
tiler = CanvasTiler(image_path="path/to/image.png", tile_width=640, tile_height=640)

# Save the tiles to an output path
tiler.saveTiles(output_path="path/to/output")

# Join the tiles from an input path and save the result
tiler.joinTiles(tiles_path="path/to/tiles", output_path="path/to/output")

License

This project is licensed under the GNU General Public License v3.0. See the LICENSE file for details.

Author

Giovanni Cascione - [e-mail]