def update_data(string):
if string:
mask = parse_jsonstring(string,
io.imread(filename, as_gray=True).shape)
else:
raise PreventUpdate
return array_to_data_url((255 * mask).astype(np.uint8))
def maskFromData(string, data, mask_filepath, image_width, image_height,
canvas_width, rect_fill):
"""
Parse the json object file and get the mask from it, then return it
Parameters :
string: data in string format
data: object that holds the canvas variables
mask_filepath: file path to save the mask to
image_width: width of the image
image_height: height of the image
canvas_width: width of the canvas
rect_fill: boolean denoting whether or not to fill rectangle objects in mask
"""
mask = np.zeros((image_height, image_width))
# get rectangular mask objects and fill them in the mask,
# otherwise if parsed it's just the outline
if 'fill' in rect_fill:
for object in data['objects']:
if object["type"] == "rect":
left = int(object["left"] * image_width / canvas_width)
top = int(object["top"] * image_width / canvas_width)
width = int(object["width"] * image_width / canvas_width)
height = int(object["height"] * image_width / canvas_width)
rect_mask = getMask(image_width, image_height, left, top,
width, height)
mask[rect_mask == 0] = 1
# parse the rest of the mask objects
mask[parse_jsonstring(string, (image_height, image_width)) == 1] = 1
plt.imsave(os.getcwd() + mask_filepath, mask, cmap=cm.gray)
def remove_background(json_data, image):
if json_data:
# ➊ imageが値をもつ場合、それを基に画像のnumpy.ndarrayに変換する
if image:
image_array = image_string_to_PILImage(image)
image_array = np.asarray(image_array)
# imageが値をもたない場合、imreadで画像を読み込む
else:
image_array = io.imread(image_path)
# ➋ 画像のアレイのサイズを変数shapeに代入する
shape = image_array.shape[:2]
# ➌ 書き込みのJSONデータをパースし、ブール値に変換する
try:
mask = parse_jsonstring(json_data, shape=shape)
except IndexError:
raise PreventUpdate
if mask.sum() > 0: # ➍
seg = superpixel_color_segmentation(image_array, mask)
else:
seg = np.ones(shape)
filled_image = np.copy(image_array)
filled_image[np.logical_not(seg)] = np.array([255, 255, 255],
dtype="uint8") # ➎
return array_to_data_url(filled_image) # ➏
else:
PreventUpdate
def update_segmentation(toggle, string, s, h, w, children, mode):
if len(children) == 0:
labs = labels
else:
labs = np.asarray(children)
mask = parse_jsonstring(string, shape=(height, width))
new_labels = modify_segmentation(labs, mask, img=img, mode=mode)
return new_labels
def segmentation(string):
if string:
mask = parse_jsonstring(string, img.shape)
seg = watershed_segmentation(img, mask)
src = color.label2rgb(seg, image=img)
else:
raise PreventUpdate
return array_to_data_url(img_as_ubyte(src))
def segmentation(string):
if string:
mask = parse_jsonstring(string,
io.imread(filename, as_gray=True).shape)
seg = watershed_segmentation(io.imread(filename, as_gray=True), mask)
src = color.label2rgb(seg, image=io.imread(filename, as_gray=True))
else:
raise PreventUpdate
return array_to_data_url(img_as_ubyte(src))
def update_data(string):
pprint(string)
if string:
mask = parse_jsonstring(string, (canvas_width, canvas_height))
else:
raise PreventUpdate
return array_to_data_url((255 * mask).astype(np.uint8))
def segmentation(string, content, NDVIcontent):
global ROIRGB, ROINDVI
if string:
data = content.encode("utf8").split(b";base64,")[1]
NDVIdata = NDVIcontent.encode("utf8").split(b";base64,")[1]
img = io.BytesIO()
imgNDVI = io.BytesIO()
img.write(base64.b64decode(data))
imgNDVI.write(base64.b64decode(NDVIdata))
img.seek(0)
imgNDVI.seek(0)
i = np.asarray(bytearray(img.read()), dtype=np.uint8)
i = cv2.imdecode(i, cv2.IMREAD_COLOR)
iNDVI = np.asarray(bytearray(imgNDVI.read()), dtype=np.uint8)
iNDVI = cv2.imdecode(iNDVI, cv2.IMREAD_COLOR)
mask = parse_jsonstring(string, (i.shape[0],i.shape[1]))
ret,thresh = cv2.threshold(np.array(mask, dtype=np.uint8), 0, 255, cv2.THRESH_BINARY)
im_floodfill = thresh.copy()
h, w = thresh.shape[:2]
m = np.zeros((h+2, w+2), np.uint8)
cv2.floodFill(im_floodfill, m, (0,0), 255)
im_floodfill_inv = cv2.bitwise_not(im_floodfill)
im_out = thresh | im_floodfill_inv
RGBimg = cv2.bitwise_and(i, i, mask=im_out)
RGBimg = cv2.cvtColor(RGBimg, cv2.COLOR_BGR2RGB)
target_size = (RGBimg.shape[1],(RGBimg.shape[0]))
iNDVI = cv2.resize(iNDVI, target_size)
NDVIimg = cv2.bitwise_and(iNDVI, iNDVI, mask=im_out)
NDVIimg = cv2.cvtColor(NDVIimg, cv2.COLOR_BGR2RGB)
contours = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[0]
cnts = sorted(contours, key=lambda c: cv2.contourArea(c), reverse=True) # finds the largest selection, which is a limitation to multiple selection (will be changed in the future versions)
(x,y,w,h) = cv2.boundingRect(cnts[0])
ROIRGB = RGBimg[y:y+h, x:x+w]
ROINDVI = NDVIimg[y:y+h, x:x+w]
cv2.imwrite(static_image_route+'RGB_cropped.png', cv2.cvtColor(ROIRGB, cv2.COLOR_RGB2BGR ))
cv2.imwrite(static_image_route+'CIR_cropped.png', cv2.cvtColor(ROINDVI, cv2.COLOR_RGB2BGR ))
with ZipFile(static_image_route+'cropped.zip', 'w') as zipObj2:
zipObj2.write(static_image_route+'RGB_cropped.png')
zipObj2.write(static_image_route+'CIR_cropped.png')
location = os.path.join(static_image_route,'cropped.zip')
else:
raise PreventUpdate
return array_to_data_url(img_as_ubyte(ROIRGB)), array_to_data_url(img_as_ubyte(ROINDVI)), html.A(html.Button('Download',style={'display':'block'
,'position':'relative',
'top': '45%','left': '45%',
'font-size': '16px',
'padding': '8px 12px',
'border-radius': '4px',
'text-align': 'center',
'align':'center',
'color':'black',
'font-family':'Times New Roman',
'textAlign':'center'}), href=location) #, html.Div([html.Button('Save Image', id='saveImg')])
def show_draw(string):
if not string:
raise PreventUpdate
# Convert image string to displayable image
mask = parse_jsonstring(string, shape)
mask = (255 * mask).astype(np.uint8)
mask = cv2.resize(mask, (400, 100), interpolation=cv2.INTER_AREA)
mask = cv2.threshold(mask, 0, 255,
cv2.THRESH_BINARY_INV | cv2.THRESH_OTSU)[1]
captcha_img = array_to_data_url(mask)
# Show image and save image string
return captcha_img, string
def classify(alteration, image):
prevent_update(image)
mask = parse_jsonstring(alteration)
with open('bug.txt', 'w') as f:
f.write(str(mask))
image2 = array_to_data_url((255 * mask).astype(np.uint8))
if alteration:
return str(
google_classify(
combine(convert_base46(image), convert_base46(image2))))
return str(google_classify(convert_base46(image)))
def update_figure_upload(string, image, algorithm):
print("update figure")
if string:
if image is None:
im = img
image = img
else:
im = image_string_to_PILImage(image)
im = np.asarray(im)
shape = im.shape[:2]
mask = parse_jsonstring(string, shape=shape)
if mask.sum() > 0:
seg = segmentation_generic(im, mask, mode=algorithm)
else:
seg = np.zeros(shape)
return image_with_contour(im, seg, shape=shape)
else:
raise PreventUpdate
def update_figure_upload(string, image):
if string:
if image is None:
im = img_app3
else:
im = image_string_to_PILImage(image)
im = np.asarray(im)
shape = im.shape[:2]
try:
mask = parse_jsonstring(string, shape=shape)
except IndexError:
raise PreventUpdate
if mask.sum() > 0:
seg = superpixel_color_segmentation(im, mask)
else:
seg = np.ones(shape)
fill_value = 255 * np.ones(3, dtype=np.uint8)
dat = np.copy(im)
dat[np.logical_not(seg)] = fill_value
return array_to_data_url(dat)
else:
raise PreventUpdate
def update_data(string):
if string:
try:
mask = parse_jsonstring(string,
shape=(canvas_height, canvas_width))
except:
return "Out of Bounding Box, click clear button and try again"
# np.savetxt('data.csv', mask) use this to save the canvas annotations as a numpy array
# Invert True and False
mask = (~mask.astype(bool)).astype(int)
image_string = array_to_data_url((255 * mask).astype(np.uint8))
# this is from canvas.utils.image_string_to_PILImage(image_string)
img = Image.open(BytesIO(base64.b64decode(image_string[22:])))
text = "{}".format(
pytesseract.image_to_string(img, lang="eng", config="--psm 6"))
return text
else:
raise PreventUpdate
def segmentation(string, content, NDVIcontent):
if string:
data = content.encode("utf8").split(b";base64,")[1]
NDVIdata = NDVIcontent.encode("utf8").split(b";base64,")[1]
img = io.BytesIO()
imgNDVI = io.BytesIO()
img.write(base64.b64decode(data))
imgNDVI.write(base64.b64decode(NDVIdata))
img.seek(0)
imgNDVI.seek(0)
i = np.asarray(bytearray(img.read()), dtype=np.uint8)
i = cv2.imdecode(i, cv2.IMREAD_COLOR)
iNDVI = np.asarray(bytearray(imgNDVI.read()), dtype=np.uint8)
iNDVI = cv2.imdecode(iNDVI, cv2.IMREAD_COLOR)
mask = parse_jsonstring(string, (i.shape[0],i.shape[1]))
ret,thresh = cv2.threshold(np.array(mask, dtype=np.uint8), 0, 255, cv2.THRESH_BINARY)
im_floodfill = thresh.copy()
h, w = thresh.shape[:2]
m = np.zeros((h+2, w+2), np.uint8)
cv2.floodFill(im_floodfill, m, (0,0), 255)
im_floodfill_inv = cv2.bitwise_not(im_floodfill)
im_out = thresh | im_floodfill_inv
RGBimg = cv2.bitwise_and(i, i, mask=im_out)
RGBimg = cv2.cvtColor(RGBimg, cv2.COLOR_BGR2RGB)
target_size = (RGBimg.shape[1],(RGBimg.shape[0]))
iNDVI = cv2.resize(iNDVI, target_size)
NDVIimg = cv2.bitwise_and(iNDVI, iNDVI, mask=im_out)
NDVIimg = cv2.cvtColor(NDVIimg, cv2.COLOR_BGR2RGB)
contours = cv2.findContours(thresh, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[0]
cnts = sorted(contours, key=lambda c: cv2.contourArea(c), reverse=True) # finds the largest selection, which is a limitation to multiple selection (will be changed in the future versions)
(x,y,w,h) = cv2.boundingRect(cnts[0])
ROIRGB = RGBimg[y:y+h, x:x+w]
ROINDVI = NDVIimg[y:y+h, x:x+w]
else:
raise PreventUpdate
return array_to_data_url(img_as_ubyte(ROIRGB)), array_to_data_url(img_as_ubyte(ROINDVI))
def update_data(string):
if string:
mask = parse_jsonstring(string, shape)
else:
raise PreventUpdate
return array_to_data_url((255 * mask).astype(np.uint8))
from dash_canvas.utils import array_to_data_url, parse_jsonstring
#from src.predict import *
shape = (200, 800)
chars = '0123456789' + string.ascii_lowercase
model = keras.models.load_model('models/4cnn_draw_model.h5')
print('receiving string with length:', len(sys.argv[1]))
# Get image string from app.py
string = sys.argv[1]
# Convert image string to numpy array
mask = parse_jsonstring(string, shape)
img = (255 * mask).astype(np.uint8)
img = cv2.resize(img, (200, 50), interpolation=cv2.INTER_AREA)
# Use model to predict image
img = img / 255.0
onehotpred = np.array(model.predict(img.reshape(1, 50, 200, 1))).reshape(5, 36)
pred = ''
# Find prediction for each character output
for i in onehotpred:
c = chars[np.argmax(i)]
pred += c
# Return prediction on app.py
print(pred)
