Spaces:
Sleeping
Sleeping
Update utils.py
Browse files
utils.py
CHANGED
@@ -9,16 +9,14 @@ incorrect = cv2.imread('wrong.png')
|
|
9 |
incorrect = cv2.cvtColor(incorrect, cv2.COLOR_BGR2RGB)
|
10 |
|
11 |
def draw_rounded_rect(img, rect_start, rect_end, corner_width, box_color):
|
12 |
-
|
13 |
-
|
14 |
-
|
15 |
-
img: The image to draw on.
|
16 |
-
rect_start: The top-left corner of the rectangle as a tuple (x1, y1).
|
17 |
-
rect_end: The bottom-right corner of the rectangle as a tuple (x2, y2).
|
18 |
-
corner_width: The width of the rounded corners.
|
19 |
-
box_color: The color of the rectangle in BGR format.
|
20 |
-
"""
|
21 |
-
|
22 |
|
23 |
x1, y1 = rect_start
|
24 |
x2, y2 = rect_end
|
@@ -47,16 +45,16 @@ def draw_rounded_rect(img, rect_start, rect_end, corner_width, box_color):
|
|
47 |
return img
|
48 |
|
49 |
def draw_dotted_line(frame, lm_coord, start, end, line_color):
|
50 |
-
|
51 |
-
This function draws a dotted line on a frame based on landmark coordinates.
|
52 |
|
53 |
-
Args:
|
54 |
-
frame: The image to draw on.
|
55 |
-
lm_coord: The landmark coordinates as a NumPy array.
|
56 |
-
start: The index of the starting landmark in the lm_coord array.
|
57 |
-
end: The index of the ending landmark in the lm_coord array.
|
58 |
-
line_color: The color of the line in BGR format.
|
59 |
-
|
60 |
|
61 |
pix_step = 0
|
62 |
|
@@ -80,25 +78,25 @@ def draw_text(
|
|
80 |
overlay_image = False,
|
81 |
overlay_type = None
|
82 |
):
|
83 |
-
|
84 |
-
This function draws text with a customizable background box on an image.
|
85 |
-
|
86 |
-
Args:
|
87 |
-
img: The image to draw on.
|
88 |
-
msg: The message to display as a string.
|
89 |
-
width: The thickness of the background box border (default: 7).
|
90 |
-
font: The font style for the text (default: cv2.FONT_HERSHEY_SIMPLEX).
|
91 |
-
pos: The top-left corner coordinates of the text box (default: (0, 0)).
|
92 |
-
font_scale: The scaling factor for the font size (default: 1).
|
93 |
-
font_thickness: The thickness of the text (default: 2).
|
94 |
-
text_color: The color of the text in BGR format (default: green - (0, 255, 0)).
|
95 |
-
text_color_bg: The color of the background box in BGR format (default: black - (0, 0, 0)).
|
96 |
-
box_offset: The offset for the background box relative to the text (default: (20, 10)).
|
97 |
-
overlay_image: Flag to display an overlay image inside the box (default: False).
|
98 |
-
overlay_type: Type of overlay image ("correct" or "incorrect") - used when overlay_image is True.
|
99 |
-
Returns:
|
100 |
-
The size of the drawn text (width, height) as a NumPy array.
|
101 |
-
|
102 |
|
103 |
offset = box_offset
|
104 |
x, y = pos
|
@@ -150,17 +148,17 @@ def draw_text(
|
|
150 |
return text_size
|
151 |
|
152 |
def find_angle(p1, p2, ref_pt = np.array([0,0])):
|
153 |
-
"""
|
154 |
-
This function calculates the angle between two points relative to a reference point.
|
155 |
|
156 |
-
|
157 |
-
|
158 |
-
|
159 |
-
|
|
|
|
|
|
|
|
|
|
|
160 |
|
161 |
-
Returns:
|
162 |
-
The angle between the two points in degrees (int).
|
163 |
-
"""
|
164 |
# Subtract the reference point from both points for normalization
|
165 |
p1_ref = p1 - ref_pt
|
166 |
p2_ref = p2 - ref_pt
|
@@ -176,18 +174,18 @@ def find_angle(p1, p2, ref_pt = np.array([0,0])):
|
|
176 |
return int(degree)
|
177 |
|
178 |
def get_landmark_array(pose_landmark, key, frame_width, frame_height):
|
179 |
-
"""
|
180 |
-
This function extracts the normalized image coordinates for a landmark.
|
181 |
|
182 |
-
|
183 |
-
|
184 |
-
|
185 |
-
|
186 |
-
|
|
|
|
|
|
|
|
|
|
|
187 |
|
188 |
-
Returns:
|
189 |
-
A NumPy array containing the normalized x and y coordinates of the landmark.
|
190 |
-
"""
|
191 |
|
192 |
denorm_x = int(pose_landmark[key].x * frame_width)
|
193 |
denorm_y = int(pose_landmark[key].y * frame_height)
|
@@ -195,20 +193,18 @@ def get_landmark_array(pose_landmark, key, frame_width, frame_height):
|
|
195 |
return np.array([denorm_x, denorm_y])
|
196 |
|
197 |
def get_landmark_features(kp_results, dict_features, feature, frame_width, frame_height):
|
198 |
-
"""
|
199 |
-
This function extracts landmark coordinates for various body parts based on a feature name.
|
200 |
|
201 |
-
|
202 |
-
kp_results: The MediaPipe pose landmark results object.
|
203 |
-
dict_features: A dictionary containing landmark key names for different body parts.
|
204 |
-
feature: The name of the body part feature to extract (e.g., 'nose', 'left', 'right').
|
205 |
-
frame_width: The width of the image frame.
|
206 |
-
frame_height: The height of the image frame.
|
207 |
|
208 |
-
|
209 |
-
|
210 |
-
|
|
|
|
|
|
|
211 |
|
|
|
|
|
212 |
|
213 |
if feature == 'nose':
|
214 |
return get_landmark_array(kp_results, dict_features[feature], frame_width, frame_height)
|
|
|
9 |
incorrect = cv2.cvtColor(incorrect, cv2.COLOR_BGR2RGB)
|
10 |
|
11 |
def draw_rounded_rect(img, rect_start, rect_end, corner_width, box_color):
|
12 |
+
|
13 |
+
# This function draws a rectangle with rounded corners on an image.
|
14 |
+
# Args:
|
15 |
+
#img: The image to draw on.
|
16 |
+
#rect_start: The top-left corner of the rectangle as a tuple (x1, y1).
|
17 |
+
#rect_end: The bottom-right corner of the rectangle as a tuple (x2, y2).
|
18 |
+
#corner_width: The width of the rounded corners.
|
19 |
+
#box_color: The color of the rectangle in BGR format.
|
|
|
|
|
20 |
|
21 |
x1, y1 = rect_start
|
22 |
x2, y2 = rect_end
|
|
|
45 |
return img
|
46 |
|
47 |
def draw_dotted_line(frame, lm_coord, start, end, line_color):
|
48 |
+
|
49 |
+
#This function draws a dotted line on a frame based on landmark coordinates.
|
50 |
|
51 |
+
#Args:
|
52 |
+
#frame: The image to draw on.
|
53 |
+
#lm_coord: The landmark coordinates as a NumPy array.
|
54 |
+
#start: The index of the starting landmark in the lm_coord array.
|
55 |
+
#end: The index of the ending landmark in the lm_coord array.
|
56 |
+
#line_color: The color of the line in BGR format.
|
57 |
+
|
58 |
|
59 |
pix_step = 0
|
60 |
|
|
|
78 |
overlay_image = False,
|
79 |
overlay_type = None
|
80 |
):
|
81 |
+
|
82 |
+
#This function draws text with a customizable background box on an image.
|
83 |
+
|
84 |
+
#Args:
|
85 |
+
#img: The image to draw on.
|
86 |
+
#msg: The message to display as a string.
|
87 |
+
#width: The thickness of the background box border (default: 7).
|
88 |
+
#font: The font style for the text (default: cv2.FONT_HERSHEY_SIMPLEX).
|
89 |
+
#pos: The top-left corner coordinates of the text box (default: (0, 0)).
|
90 |
+
#font_scale: The scaling factor for the font size (default: 1).
|
91 |
+
#font_thickness: The thickness of the text (default: 2).
|
92 |
+
#text_color: The color of the text in BGR format (default: green - (0, 255, 0)).
|
93 |
+
#text_color_bg: The color of the background box in BGR format (default: black - (0, 0, 0)).
|
94 |
+
#box_offset: The offset for the background box relative to the text (default: (20, 10)).
|
95 |
+
#overlay_image: Flag to display an overlay image inside the box (default: False).
|
96 |
+
#overlay_type: Type of overlay image ("correct" or "incorrect") - used when overlay_image is True.
|
97 |
+
#Returns:
|
98 |
+
#The size of the drawn text (width, height) as a NumPy array.
|
99 |
+
|
100 |
|
101 |
offset = box_offset
|
102 |
x, y = pos
|
|
|
148 |
return text_size
|
149 |
|
150 |
def find_angle(p1, p2, ref_pt = np.array([0,0])):
|
|
|
|
|
151 |
|
152 |
+
#This function calculates the angle between two points relative to a reference point.
|
153 |
+
|
154 |
+
#Args:
|
155 |
+
#p1: The first point coordinates as a NumPy array (x, y).
|
156 |
+
#p2: The second point coordinates as a NumPy array (x, y).
|
157 |
+
#ref_pt: The reference point coordinates as a NumPy array (default: [0, 0]).
|
158 |
+
|
159 |
+
#Returns:
|
160 |
+
#The angle between the two points in degrees (int).
|
161 |
|
|
|
|
|
|
|
162 |
# Subtract the reference point from both points for normalization
|
163 |
p1_ref = p1 - ref_pt
|
164 |
p2_ref = p2 - ref_pt
|
|
|
174 |
return int(degree)
|
175 |
|
176 |
def get_landmark_array(pose_landmark, key, frame_width, frame_height):
|
|
|
|
|
177 |
|
178 |
+
#This function extracts the normalized image coordinates for a landmark.
|
179 |
+
|
180 |
+
#Args:
|
181 |
+
#pose_landmark: A MediaPipe pose landmark object.
|
182 |
+
#key: The key name of the landmark to extract (e.g., 'nose', 'shoulder.x').
|
183 |
+
#frame_width: The width of the image frame.
|
184 |
+
#frame_height: The height of the image frame.
|
185 |
+
|
186 |
+
#Returns:
|
187 |
+
#A NumPy array containing the normalized x and y coordinates of the landmark.
|
188 |
|
|
|
|
|
|
|
189 |
|
190 |
denorm_x = int(pose_landmark[key].x * frame_width)
|
191 |
denorm_y = int(pose_landmark[key].y * frame_height)
|
|
|
193 |
return np.array([denorm_x, denorm_y])
|
194 |
|
195 |
def get_landmark_features(kp_results, dict_features, feature, frame_width, frame_height):
|
|
|
|
|
196 |
|
197 |
+
#This function extracts landmark coordinates for various body parts based on a feature name.
|
|
|
|
|
|
|
|
|
|
|
198 |
|
199 |
+
#Args:
|
200 |
+
#kp_results: The MediaPipe pose landmark results object.
|
201 |
+
#dict_features: A dictionary containing landmark key names for different body parts.
|
202 |
+
#feature: The name of the body part feature to extract (e.g., 'nose', 'left', 'right').
|
203 |
+
#frame_width: The width of the image frame.
|
204 |
+
#frame_height: The height of the image frame.
|
205 |
|
206 |
+
#Returns:
|
207 |
+
#A list containing the landmark coordinates (as NumPy arrays) or raises an error if the feature is invalid.
|
208 |
|
209 |
if feature == 'nose':
|
210 |
return get_landmark_array(kp_results, dict_features[feature], frame_width, frame_height)
|