def capture_setting():
# intialising the key information
counter = 1
marker = 1
distance = 0
predicted_character_list = []
predicted_color_list = []
end = time.time()
start = time.time()
config = Settings()
save = Saving(config.name_of_folder, config.exist_ok)
if config.capture == "pc":
if config.testing == "video":
cap = cv2.VideoCapture(config.media)
else:
cap = cv2.VideoCapture(0)
cap.set(cv2.CAP_PROP_FRAME_WIDTH, 960) # 800 default
cap.set(3, 960) # 800 default
cap.set(4, 540) # 800 default
cap.set(cv2.CAP_PROP_FRAME_HEIGHT, 540)
cap.set(cv2.CAP_PROP_FPS, 60)
time.sleep(2) # allows the camera to start-up
print('Camera on')
while True:
if counter == 1:
if config.pause:
distance = input("Are you Ready?")
if counter == 1 or end - start < 10:
end = time.time()
ret, frame = cap.read()
if config.Step_camera:
cv2.imshow('frame', frame)
k = cv2.waitKey(5) & 0xFF
if k == 27:
break
color, roi, frame, success = detection(frame, config)
if success:
counter = counter + 1
# time that the target has been last seen
start = time.time()
predicted_character, contour_image, chosen_image = character_recognition.character(color)
predicted_color, processed_image = colour_recognition.colour(color)
predicted_character_list.append(predicted_character)
predicted_color_list.append(predicted_color)
if config.save_results:
name_of_results = ["color", "roi", "frame","contour_image","processed_image", "chosen_image"]
image_results = [color, roi, frame, contour_image, processed_image, chosen_image]
for value, data in enumerate(name_of_results):
image_name = f"{marker}_{data}_{counter}.jpg"
image = image_results[value]
if image is not None:
save.save_the_image(image_name, image)
if counter == 8:
print("Starting Recognition Thread")
common_character = Counter(predicted_character_list).most_common(1)[0][0]
common_color = Counter(predicted_color_list).most_common(1)[0][0]
solution(counter, marker, common_character, common_color, save.save_dir)
predicted_character_list = []
predicted_color_list = []
else:
print("Starting Recognition Thread")
common_character = Counter(predicted_character_list).most_common(1)[0][0]
common_color = Counter(predicted_color_list).most_common(1)[0][0]
solution(counter, marker, common_character, common_color, save.save_dir)
predicted_character_list = []
predicted_color_list = []
elif config.capture == "pi":
from picamera.array import PiRGBArray
from picamera import PiCamera
camera = PiCamera()
camera.resolution = (1280, 720)
camera.brightness = 50 # 50 is default
camera.framerate = 90
camera.awb_mode = 'auto'
camera.shutter_speed = camera.exposure_speed
cap = PiRGBArray(camera, size=(1280, 720))
for image in camera.capture_continuous(cap, format="bgr", use_video_port=True):
# to start the progress of capture and don't stop unless the counter increases and has surpass 5 seconds
if counter == 1 or end - start < 10:
frame = image.array
end = time.time()
color, roi, frame, success = detection(frame, config)
if success:
counter = counter + 1
# time that the target has been last seen
start = time.time()
predicted_character, contour_image, chosen_image = character_recognition.character(color)
predicted_color, processed_image = colour_recognition.colour(color)
predicted_character_list.append(predicted_character)
predicted_color_list.append(predicted_color)
if config.save_results:
name_of_results = ["color", "roi", "frame","contour_image","processed_image", "chosen_image"]
image_results = [color, roi, frame, contour_image, processed_image, chosen_image]
for value, data in enumerate(name_of_results):
image_name = f"{marker}_{data}_{counter}.jpg"
image = image_results[value]
if image is not None:
save.save_the_image(image_name, image)
if counter == 8:
print("Starting Recognition Thread")
common_character = Counter(predicted_character_list).most_common(1)[0][0]
common_color = Counter(predicted_color_list).most_common(1)[0][0]
marker, counter = solution(counter, marker, common_character, common_color, save.save_dir)
predicted_character_list = []
predicted_color_list = []
else:
print("Starting Recognition Thread")
common_character = Counter(predicted_character_list).most_common(1)[0][0]
common_color = Counter(predicted_color_list).most_common(1)[0][0]
marker, counter = solution(counter, marker, common_character, common_color, save.save_dir)
predicted_character_list = []
predicted_color_list = []
cap.truncate(0)
elif config.capture == "image":
cap = [] # to store the names of the images
data_dir = Path(config.media)
# the following code interite over the extension that exist within a folder and place them into a single list
image_count = list(itertools.chain.from_iterable(data_dir.glob(pattern) for pattern in ('*.jpg', '*.png')))
# image_count = len(list(data_dir.glob('*.jpg')))
for name in image_count:
# head, tail = ntpath.split(name)
filename = Path(name) # .stem removes the extension and .name grabs the filename with extension
cap.append(filename)
test_image = cv2.imread(str(filename))
marker = Path(name).stem # grabs the name with the extension
color, roi, frame, success = detection(test_image, config)
if success:
predicted_character, contour_image, chosen_image = character_recognition.character(color)
predicted_color, processed_image = colour_recognition.colour(color)
_, _ = solution(counter, marker, predicted_character, predicted_color, save.save_dir)
if config.save_results:
name_of_results = ["color", "roi", "frame","contour_image","processed_image", "chosen_image", color, roi, frame, contour_image, processed_image, chosen_image]
for value in range(5):
image_name = f"{marker}_{name_of_results[value]}.jpg"
image = name_of_results[value + 6]
if image is not None:
save.save_the_image(image_name, image)
print("Detected and saved a target")
print(f"there is a total image count of {len(image_count)} and frames appended {len(cap)}")
