[UFOCR-17] Add more image processing algorithms

[zamazan4ik]

List of required algorithms:

• Deblurring (Wiener filter, Gaussian filter, smth else). Now we have nothing.
• Cropping/Autocropping (Canny-based, Hougline-based, feature-detection-based,etc.). Now we have unworkable crop. Autocropping is not implemented yet.
• Binarization(Otsu, COCOCLUST and a lot other). Now we have a lot of handwriiten algorithm. Should be rewritten as much as possible with OpenCV.
• Invert. Now we have nothing.
• Deskew (probably Leptonica-based). Manual deskew is implemented, auto deskew does not work correctly.
• Remove distortions (Trapezoidal, etc). Now we have nothing.
• Denoising (Bilateral filter, adaptive Bilateral filter, etc). Now we have simple bilateral filter, should be improved a lot.
• Remove lines. Now we have nothing.
• Auto choosing algorithm for binarization (probably should be implemented with neural networks). Now we have nothing.

[zamazan4ik]

Added initial support for: deblurring, blur detection, OpenCV-based binarising(OTSU).