python - Image loses quality with cv2.warpPerspective -

i working opencv 3.1 , python.

my problem comes when try deskew (fix tilt of) image text. using cv2.warpperspective make possible, image loses lot of quality. can see here original part of image:

and then, here, "rotated" image:

it smoothed.

i using morphological transformation like:

kernel = np.ones((2, 2), np.uint8) blur_image = cv2.erode(tresh, kernel, iterations=1) 

and

white_mask2 = cv2.morphologyex(white_mask2, cv2.morph_open, kernel) 

to see if improves something, nothing.

i saw this example here in so, guys had same problem: ,

so, don't have idea can do. maybe there way not losing quality of image, or, there's method rotate image without quality lost. know method:

root_mat = cv2.getrotationmatrix2d(to_rotate_center, angle, 1.0) result = cv2.warpaffine(to_rotate, root_mat, to_rotate.shape, flags=cv2.inter_linear) 

but doesn't work me, have rotate every rectangle here:

and not whole image. means, best way found it, warpperspective, , works fine, quality loss. appreciate advice avoid quality lost.

the problem related interpolation required warping. if don't want things appear smoother, should switch default interpolation method, inter_linear one, such inter_nearest. sharpness of edges.

try flags=cv2.inter_nearest on warp call, warpaffine() or warpperspective().

interpolation flags listed here.

enum interpolationflags {      inter_nearest = 0,      inter_linear = 1,      inter_cubic = 2,      inter_area = 3,      inter_lanczos4 = 4,      inter_max = 7,      warp_fill_outliers = 8,      warp_inverse_map = 16  }