def uploadUGProductImage(request):
if request.method == 'POST':
process_image(request.POST["product1_url"], request.POST["product1_filename"])
process_image(request.POST["product2_url"], request.POST["product2_filename"])
# use process_images.py
return HttpResponse("looks like it worked")
return HttpResponse("Not a post request")
def main(cards_file, debug=False):
with open(cards_file) as f:
content = f.readlines()
card_list = [line.rstrip('\n') for line in content]
download_images(card_list)
for card_name in card_list:
print(card_name)
process_image(card_name, debug)
def predict(image_path, category_names, model, top_k, device):
device = torch.device(
"cuda:0" if torch.cuda.is_available() and device == 'gpu' else 'cpu')
image = process_image(image_path)
tensor_im = torch.from_numpy(image)
tensor_im = tensor_im.permute(2, 0, 1).type(torch.cuda.FloatTensor)
tensor_im = tensor_im.unsqueeze(0)
with torch.no_grad():
model.to(device)
tensor_im.to(device)
output = model.forward(tensor_im)
probs = torch.exp(output)
probabilities, idx = torch.topk(probs, top_k)
idx = idx.cpu().numpy()[0]
probabilities = probabilities.cpu().numpy()[0]
classes = {val: key for key, val in model.class_to_idx.items()}
top_classes = [classes[each] for each in idx]
with open(category_names, 'r') as f:
category_names = json.load(f)
label_idx = re.split('/', image_path)
label_idx = label_idx[2]
label = category_names[label_idx]
predicted_labels = [category_names[i] for i in top_classes]
round_prob = list(np.around(np.array(probabilities), 2))
predict_dict = dict(zip(predicted_labels, round_prob))
print('Predicted Image Label: ', label)
print('Top ' + str(top_k) + ' Probabilities')
for k, v in predict_dict.items():
print(str(k) + ': ' + str(v))
def main(img):
#返回车牌定位
cropImg = image_position(img)
#车牌倾斜矫正
rotated = correct_image(cropImg)
#分割字符
images = process_image(rotated)
#预测
prediction(images)
def make_voc(imlist):
featlist = [ process_image(im) for im in imlist ]
voc = Vocabulary(VOC_NAME)
voc.train(featlist, 1000, 10)
# saving vocabulary
with open(VOC_FILE, 'wb') as f:
cPickle.dump(voc, f)
print 'vocabulary', voc.name, 'has', voc.nbr_words, 'words'
def unify_test_function(model, batch):
if len(batch) == 3:
x, y, _ = batch
else:
x, y = batch
x, y = x.to(model.device), y.to(model.device)
pred = process_image(model, x, model.input_size, model.n_channels)
loss = F.mse_loss(pred.detach(), y)
pcc = pearson_corrcoef(pred.reshape(-1), y.reshape(-1))
return pred.detach(), loss.detach(), pcc.detach()
def telemetry(sid, data):
if data:
# The current steering angle of the car
steering_angle = data["steering_angle"]
# The current throttle of the car
throttle = data["throttle"]
# The current speed of the car
speed = data["speed"]
# The current image from the center camera of the car
imgString = data["image"]
image = Image.open(BytesIO(base64.b64decode(imgString)))
image_array = np.asarray(image)
image_4d = image_array[None, :, :, :]
# Pre-process
image_3d = process_image(convert_4d_to_3d(image_4d))
image_4d = convert_3d_to_4d(image_3d)
steering_angle = float(model.predict(image_4d, batch_size=1))
#throttle = controller.update(float(speed))
throttle = 0.2
print(steering_angle, throttle)
send_control(steering_angle, throttle)
# save frame
if args.image_folder != '':
timestamp = datetime.utcnow().strftime('%Y_%m_%d_%H_%M_%S_%f')[:-3]
image_filename = os.path.join(args.image_folder, timestamp)
image.save('{}.jpg'.format(image_filename))
else:
# NOTE: DON'T EDIT THIS.
sio.emit('manual', data={}, skip_sid=True)
def test_process_image():
assert pi.process_image(TEST_IMAGE) != None
def pipe_images():
# if not request.json or not 'data' in request.json:
# abort(400)
print(request.json)
result = cf.process_image(request.json.get("data"))
return jsonify({"status": result[0], "name": result[1]}), 201