cv-analysis-service/cv_analysis/layout_parsing.py

78 lines
2.3 KiB
Python

from itertools import compress
from itertools import starmap
from operator import __and__
import cv2
import numpy as np
from cv_analysis.utils.structures import Rectangle
from cv_analysis.utils.post_processing import (
remove_overlapping,
remove_included,
has_no_parent,
)
from cv_analysis.utils.visual_logging import vizlogger
def is_likely_segment(rect, min_area=100):
return cv2.contourArea(rect, False) > min_area
def find_segments(image):
contours, hierarchies = cv2.findContours(image, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)
mask1 = map(is_likely_segment, contours)
mask2 = map(has_no_parent, hierarchies[0])
mask = starmap(__and__, zip(mask1, mask2))
contours = compress(contours, mask)
rectangles = (cv2.boundingRect(c) for c in contours)
return rectangles
def parse_layout(image: np.array):
image = image.copy()
image_ = image.copy()
if len(image_.shape) > 2:
image_ = cv2.cvtColor(image_, cv2.COLOR_BGR2GRAY)
vizlogger.debug(image_, "layout01_start.png")
image_ = cv2.GaussianBlur(image_, (7, 7), 0)
vizlogger.debug(image_, "layout02_blur.png")
thresh = cv2.threshold(image_, 0, 255, cv2.THRESH_BINARY_INV + cv2.THRESH_OTSU)[1]
vizlogger.debug(image_, "layout03_theshold.png")
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5, 5))
vizlogger.debug(kernel, "layout04_kernel.png")
dilate = cv2.dilate(thresh, kernel, iterations=4)
vizlogger.debug(dilate, "layout05_dilate.png")
rects = list(find_segments(dilate))
# -> Run meta detection on the previous detections TODO: refactor
for rect in rects:
x, y, w, h = rect
cv2.rectangle(image, (x, y), (x + w, y + h), (0, 0, 0), -1)
cv2.rectangle(image, (x, y), (x + w, y + h), (255, 255, 255), 7)
vizlogger.debug(image, "layout06_rectangles.png")
_, image = cv2.threshold(image, 254, 255, cv2.THRESH_BINARY)
vizlogger.debug(image, "layout07_threshold.png")
image = ~image
vizlogger.debug(image, "layout08_inverse.png")
if len(image.shape) > 2:
image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
vizlogger.debug(image, "layout09_convertcolor.png")
rects = find_segments(image)
# <- End of meta detection
rects = remove_included(rects)
rects = remove_overlapping(rects)
return list(map(Rectangle.from_xywh, rects))