def predict(configs, detection):
global predictor
width = configs['input_width']
height = configs['input_height']
image = read_image(configs)
input = preprocess_image(image, configs)
tensor = pm.PaddleTensor()
tensor.dtype = pm.PaddleDType.FLOAT32
tensor.shape = (1, 3, width, height)
tensor.data = pm.PaddleBuf(input)
paddle_data_feeds = [tensor]
print('prediction is running ...')
outputs = predictor.Run(paddle_data_feeds)
assert len(
outputs
) == 1, 'error numbers of tensor returned from Predictor.Run function !!!'
output = np.array(outputs[0], copy=False)
print('output:', output)
print('\nprediction result :')
print('\t nDim: ' + str(output.ndim))
print('\tShape: ' + str(output.shape))
print('\tDType: ' + str(output.dtype))
# print(output)
# print('')
if detection:
detect(output, configs)
else:
classify(output, configs)
def predict_camera(self, configs):
"""
using camera to predict image
:param configs:
:return: None
"""
# for serial communication
# camera_serial = SerialThread('serial_thread')
# camera_serial.start()
self.video_thread.start()
init_flag = True
while True:
frame_read = self.video_thread.get_image()
if frame_read is None:
print("[ERROR] fail to read frame...")
break
else:
print("[INFO] camera prediction running...")
if init_flag:
init_flag = False
continue
image = self.preprocess_image(frame_read, configs)
self.tensor.data = pm.PaddleBuf(image)
paddle_data_feeds = [self.tensor]
outputs = self.predictor.Run(paddle_data_feeds)
output = np.array(outputs[0], copy=False)
# for test
self.show_result_in_console(frame_read, output,
configs['threshold'])
def predict_video(self, configs):
"""
read a video to precdict
:param configs: detection configuration
:return: None
"""
self.video_thread.start()
init_flag = True
while True:
frame = self.video_thread.get_image()
if frame is None:
print('[INFO] fail to read frame ...')
break
if init_flag:
print('[INFO] prediction is running ...')
init_flag = False
image = self.preprocess_image(frame, configs)
self.tensor.data = pm.PaddleBuf(image)
paddle_data_feeds = [self.tensor]
outputs = self.predictor.Run(paddle_data_feeds)
assert len(
outputs
) == 1, 'error numbers of tensor returned from Predictor.Run function !!!'
output = np.array(outputs[0], copy=False)
# final
# todo: write the result per frame to new video
# for test
self.show_result_in_console(frame, output, configs['threshold'])
def predict_image(self, configs):
"""
read a image to predict
:param configs: detection configuration
:return: None
"""
image = self.read_image(configs)
input = self.preprocess_image(image, configs)
self.tensor.data = pm.PaddleBuf(input)
paddle_data_feeds = [self.tensor]
print('prediction is running ...')
outputs = self.predictor.Run(paddle_data_feeds)
assert len(
outputs
) == 1, 'error numbers of tensor returned from Predictor.Run function !!!'
output = np.array(outputs[0], copy=False)
print('\nprediction result :')
print('\t nDim: ' + str(output.ndim))
print('\tShape: ' + str(output.shape))
print('\tDType: ' + str(output.dtype))
image = self.read_image(configs)
self.draw_results(image, output, configs['threshold'])
def predict(self, input_data):
'''
PaddleMobile模型预测
参数：输入数据张量、图像数据、预测器
返回：模型预测结果
'''
self.tensor.data = pm.PaddleBuf(input_data)
paddle_data_feeds = [self.tensor]
outputs = self.predictor.Run(paddle_data_feeds)
result = np.array(outputs[0])
return result
def tensor_deal(self, origin):
tensor_img = origin.resize((256, 256), Image.BILINEAR)
if tensor_img.mode != 'RGB':
tensor_img = tensor_img.convert('RGB')
tensor_img = np.array(tensor_img).astype('float32').transpose(
(2, 0, 1))
tensor_img -= 127.5
tensor_img *= 0.007843
tensor_img = tensor_img[np.newaxis, :]
tensor = pm.PaddleTensor()
tensor.dtype = pm.PaddleDType.FLOAT32
tensor.shape = (1, 3, 256, 256)
tensor.data = pm.PaddleBuf(tensor_img)
paddle_data_feeds = [tensor]
return paddle_data_feeds
def predict(self, image):
"""
PaddleMobile模型预测
:param image: 图像数据
:return: 模型预测结果
"""
image = self.preprocess_image(image)
self.tensor.data = pm.PaddleBuf(image)
paddle_data_feeds = [self.tensor]
outputs = self.predictor.Run(paddle_data_feeds)
result = np.array(outputs[0])
# if result[0] == -1.0:
# return []
height, width, _ = image.shape
boxes = self.convert_predict_result(result, height, width)
return boxes
while True:
# 保存图像 调试用
select.select((video,), (), ())
image_data = video.read_and_queue()
frame = cv2.imdecode(np.frombuffer(image_data, dtype=np.uint8), cv2.IMREAD_COLOR)
cv2.imwrite("test.jpg", frame)
# 图像预处理
origin, img = dataset(video)
tensor_img = origin.resize((244, 244), Image.BILINEAR)
if tensor_img.mode != 'RGB':
tensor_img = tensor_img.convert('RGB')
tensor_img = np.array(tensor_img).astype('float32').transpose((2, 0, 1)) # HWC to CHW
tensor = pm.PaddleTensor()
tensor.dtype =pm.PaddleDType.FLOAT32
tensor.shape = (1,1,244,244)
tensor.data = pm.PaddleBuf(tensor_img)
paddle_data_feeds = [tensor]
# 将图像输入神经网络并获得输出
outputs = predictor.Run(paddle_data_feeds)
# 处理输出结果并返回最终预测值
output = np.array(outputs[0], copy = False)
# print(output)
number = np.argsort(output)
print("The number is ", number[0][-1])