def getImages(filename):
imagePath = filename
sidePath = cropmarkup.getMarkupTab(imagePath)
landmarks = blobdetection.getLandmarks(sidePath)
landmarks.sort(key=lambda x: x.point.pt[1])
imageArray = []
for x in range(0, len(landmarks)):
# print x, len(landmarks)
if (x+1 >= len(landmarks)):
imageArray.append(cropsection.getSection(landmarks[x].point, 0, imagePath))
else:
imageArray.append(cropsection.getSection(landmarks[x].point, landmarks[x+1].point, imagePath))
return imageArray
def getImages(imagePath):
"""
Args:
filename: the path to the imge file to segment
Returns:
Returns the sections filenames that are the results.
"""
# Read in the image
im = cv2.imread(imagePath, cv2.IMREAD_COLOR)
height, width = im.shape[:2]
print(im[5,:,2])
cv2.imshow('nothing', im)
# Requires that red is present to find appropriate point
markupImage = cropmarkup.getMarkupTab(im)
if (markupImage is None):
raise Exception("Can't find a margin.")
cv2.imwrite('Image.jpg', markupImage)
# Find all of the markup points in the image
shapes = blobdetection.getShapes(markupImage)
if (shapes == None or len(shapes) == 0 ):
raise Exception("Can't find any shapes.")
# Sort images by the ones closest to the top
shapes.sort(key=lambda x: x.origin[1])
imageArray = []
merge_close_neighbors(shapes)
# For each landmark, cut a section from either the top of the page or from landmark to landmark
for x in range(0, len(shapes)):
tempImage = None
if (x + 1 >= len(shapes)):
tempImage = cropsection.getSection(im, shapes[x].origin, (width, height))
else:
tempImage = cropsection.getSection(im,shapes[x].origin, (width, shapes[x+1].origin[1]))
tempPath = "static/temp/{}.jpg".format(uuid.uuid4())
imageArray.append(tempPath)
cv2.imwrite(tempPath, tempImage)
return imageArray
