How to get the number of closed contours?

[jkraehling]

Hi, first of all thanks for the great library!

I'm trying to get the closed contours in dxf files with simple lines, arcs and circles.

I attached a test.dxf (compressed to test.zip): test.zip
And I uploaded a picture of the file below.

So it has six closed contours.
But I don't know how to get this number.

I started to get the start and end points, now I should connect the lines and arcs, but I don't know how to continue or how to use this information to continue. I want to connect them with something like if math.isclose(sp_arc, ep_arc,abs_tol=1e-4):

`
import ezdxf
import math

drawing = ezdxf.readfile('test.dxf')
msp = drawing.modelspace()

number_cc = 0

for ei in msp:
    
    #Circle is simple, in my case they are always a closed contour
    if ei.dxftype() == 'CIRCLE':
        number_cc += 1
    
    elif ei.dxftype() == 'LINE':
        sp_line = ei.dxf.start
        ep_line = ei.dxf.end
    
    elif ei.dxftype() == 'ARC':
        sp_arc = ei.start_point
        ep_arc = ei.end_point
    
    #How to continue?
    
print('Number of closed contours: '+ str(number_cc))

`

[mozman]

This is not a DXF or ezdxf related question, but a question about a solution algorithm for a very special problem. You'd better post this question on StackOverflow, where it's much more likely to get an answer.

[melihpeker-pointr]

You can use OpenCV for finding closed contours. First draw the image and then apply CV algorithms on the output. Here, you can use:

    param = {"show_hatch": 0, "hatch_pattern": 0}
    fig = plt.figure()
    ax = fig.add_axes([0, 0, 1, 1])
    ctx = RenderContext(doc)
    ctx.current_layout.set_colors(bg="#FFFFFFFF")
    out = MatplotlibBackend(ax, params=param)
    Frontend(ctx, out).draw_layout(doc.modelspace())

    canvas = FigureCanvas(fig)
    canvas.draw()
    graph_image = np.array(fig.canvas.get_renderer()._renderer)
    # convert to gray
    imgGray = cv2.cvtColor(graph_image, cv2.COLOR_BGR2GRAY)
    # binary threshold
    img = cv2.threshold(imgGray, 127, 255, cv2.THRESH_BINARY_INV)[1]  # ensure binary
    # find contours
    contours, hierarchy = cv2.findContours(fig, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
    contours_poly = [None] * len(contours)
    print("Number of contours: " + str(len(contours)))`

For more information about algorithm, see: https://docs.opencv.org/master/d9/d8b/tutorial_py_contours_hierarchy.html