def recognize_text(self):
self.text_lines, images = [], []
for (x0, y0), (x1, y1), (x2, y2), (x3, y3) in self.text_areas['boxes']:
center = round((x0 + x1) / 2), round((y0 + y2) / 2)
region = self.bgr[round(y0):round(y2), round(x0):round(x2)]
d = pytesseract.image_to_data(region, output_type=pytesseract.Output.DICT)
words = []
for text, conf, x, y, w, h in zip(d['text'], map(float, d['conf']), d['left'], d['top'], d['width'], d['height']):
if conf > 0 and text:
words.append({
"text": text.strip(),
"conf": conf,
"center": (round(x0 + x + w / 2), round(y0 + y + h / 2))
})
if words:
line = {
"center": center,
"top": round(y0), "bottom": round(y2),
"left": round(x0), "right": round(x1),
"text": " ".join([word["text"] for word in words]),
"words": words,
"conf": sum([word["conf"] for word in words]) // len(words)
}
self.text_lines.append(line)
if self.debug:
is_success, im_buf_arr = cv2.imencode(".png", region)
byte_im = im_buf_arr.tobytes()
images.append((byte_im, line["text"]))
if self.debug:
display(create_gallery(images))
def notebook_image_marked(image, *marks, comment=""):
image = image.copy()
for center in marks:
image = cv2.circle(image, center, 20, (0, 0, 255), 3)
is_success, im_buf_arr = cv2.imencode(".png", image)
byte_im = im_buf_arr.tobytes()
display(create_gallery([(byte_im, comment)]))
def notebook_image(temp, file, comment=None):
if comment is None:
comment = file
with open(os.path.join(temp, file), 'rb') as f:
image = f.read()
display(create_gallery([(image, comment)]))
def get_child_windows_from_xid(xid):
'''
Given an X window id, return the xid's of its children
'''
try:
with display() as d:
root = d.screen().root
xw = find_xwindow_by_id(xid, root)
children = []
for c in xw.query_tree().children:
children.append(xid_to_str(c.id))
return children
except Exception:
return []
def get_child_windows_from_xid(xid):
'''
Given an X window id, return the xid's of its children
'''
try:
with display() as d:
root = d.screen().root
xw = find_xwindow_by_id(xid, root)
children = []
for c in xw.query_tree().children:
children.append(xid_to_str(c.id))
return children
except:
return []
def type_string(s, interval=0):
if interval:
interval = float(interval) / 1000
def i():
time.sleep(interval)
else:
def i():
pass
for c in s:
sym = Xlib.XK.string_to_keysym(c)
if sym:
shifted = c.isupper()
kc = display().keysym_to_keycode(sym)
if shifted:
key_down("SHIFT")
key_down(kc)
key_up(kc)
if shifted:
key_up("SHIFT")
else:
key_down("CONTROL")
key_down("SHIFT")
key_down("U")
key_up("U")
key_up("SHIFT")
key_up("CONTROL")
u = ord(c)
u = format(u, "04X")
for d in u:
key_down(d)
key_up(d)
key_down("SPACE")
key_up("SPACE")
def key_up(vk):
if isinstance(vk, str):
vk = KEY_CODES[vk.upper()]
Xlib.ext.xtest.fake_input(display(), Xlib.X.KeyRelease, vk)
display().sync()