def test_read(self):
input1 = mx.nd.random.uniform(0, 255, shape=(3, 256, 256))
input_buf1 = self._write_image(input1)
output1 = image.read(input_buf1)
assert output1.shape == (256, 256, 3), "Read method failed. Got %s shape." % (str(output1.shape))
input2 = mx.nd.random.uniform(0, 255, shape=(1, 128, 128))
input_buf2 = self._write_image(input2, flag=0)
output2 = image.read(input_buf2, flag=0)
assert output2.shape == (128, 128, 1), "Read method failed. Got %s shape." % (str(output2.shape))
def _preprocess(self, data):
img_list = []
for idx, img in enumerate(data):
input_shape = self.signature['inputs'][idx]['data_shape']
# We are assuming input shape is NCHW
[h, w] = input_shape[2:]
if input_shape[1] == 1:
img_arr = image.read(img, 0)
else:
img_arr = image.read(img)
img_arr = image.resize(img_arr, w, h)
img_arr = image.transform_shape(img_arr)
img_list.append(img_arr)
return img_list
def _preprocess(self, data):
input_shape = self.signature['inputs'][0]['data_shape']
height, width = input_shape[2:]
img_arr = image.read(data[0])
img_arr = image.resize(img_arr, width, height)
img_arr = image.color_normalize(img_arr, nd.array([127.5]),
nd.array([127.5]))
img_arr = image.transform_shape(img_arr)
return [img_arr]
def _preprocess(self, data):
img_list = []
for idx, img in enumerate(data):
input_shape = self.signature['inputs'][idx]['data_shape']
# We are assuming input shape is NCHW
[h, w] = input_shape[2:]
if input_shape[1] == 1:
img_arr = image.read(img, 0)
else:
img_arr = image.read(img)
img_arr = image.resize(img_arr, w, h)
rgb_mean = mx.nd.array([123.68, 116.779, 103.939])
rgb_std = mx.nd.array([58.395, 57.12, 57.375])
img_arr = img_arr.astype('float32')
img_arr = mx.image.color_normalize(img_arr,
mean=rgb_mean,
std=rgb_std)
img_arr = mx.nd.reshape(img_arr, (w, h, input_shape[1]))
img_arr = image.transform_shape(img_arr)
img_list.append(img_arr)
return img_list
def _preprocess(self, data):
img_list = []
for idx, img in enumerate(data):
input_shape = self.signature['inputs'][idx][
'data_shape'] # Input shape is NCHW
[h, w] = input_shape[2:]
img_arr = image.read(
img, 0) #Set flag to 0 for reading grayscale images
img_arr = image.resize(img_arr, w, h)
img_arr = image.transform_shape(img_arr)
img_list.append(img_arr)
return img_list
def preprocess(self, request):
"""
Pre-process requests by attempting to extract face image, and transforming to fit the model's input
Returns
-------
list of NDArray
Processed images in the model's expected input shape
"""
img_list = []
input_shape = self.signature['inputs'][0]['data_shape']
[height, width] = input_shape[2:]
param_name = self.signature['inputs'][0]['data_name']
# Iterate over all input images provided with the request, transform and append for inference
for idx, data in enumerate(request):
# Extract the input image
img = data.get(param_name)
if img is None:
img = data.get("body")
if img is None:
img = data.get("data")
if img is None or len(img) == 0:
self.error = "Empty image input"
return None
try:
img_arr = image.read(img).asnumpy()
except Exception as e:
logging.warning(e, exc_info=True)
self.error = "Corrupted image input"
return None
# Try to identify face to crop
face = crop_face(img_arr)
if face is not None:
face = transform.resize(face, (height, width))
# If no face identified - use the entire input image
else:
face = cv.cvtColor(img_arr, cv.COLOR_BGR2GRAY)
# Transform image into tensor of the required shape
face = np.resize(face, input_shape)
face = normalize(face, height, width)
face = mx.nd.array(face)
img_list.append(face)
return img_list
def _preprocess(self, data):
"""
Input image buffer from data is read into NDArray. Then, resized to
expected shape. Swaps axes to convert image from BGR format to RGB.
Returns the preprocessed NDArray as a list for next step, Inference.
"""
# Read input
input_image = image.read(data[0])
# Save original input image shape.
# This is required for preparing the bounding box of the detected object relative to
# original input
self.input_height = input_image.shape[0]
self.input_width = input_image.shape[1]
# Transform input image - resize, BGR to RGB.
# Reuse MXNetVisionService _preprocess to achieve above transformations.
return super(SSDService, self)._preprocess(data)
def _preprocess(self, data):
"""
Input image buffer from data is read into NDArray. Then, resized to
expected shape. Swaps axes to convert image from BGR format to RGB.
Returns the preprocessed NDArray as a list for next step, Inference.
"""
# Read input
input_image = image.read(data[0])
# Save original input image shape.
# This is required for preparing the bounding box of the detected object relative to
# original input
self.input_height = input_image.shape[0]
self.input_width = input_image.shape[1]
# Transform input image - resize, BGR to RGB.
# Reuse MXNetVisionService _preprocess to achieve above transformations.
return super(SSDService, self)._preprocess(data)
def _preprocess(self, data):
img_list = []
for idx, img in enumerate(data):
input_shape = self.signature['inputs'][idx]['data_shape']
# We are assuming input shape is NCHW
img_arr = image.read(img)
img_arr, im_scale = rcnn_resize(img_arr,
SHORTER_SIZE,
MAX_SIZE,
stride=IMAGE_STRIDE)
rgb_mean = mx.nd.array([[[0, 0, 0]]])
img_arr = img_arr.astype('float32')
img_arr = img_arr - rgb_mean
img_arr = image.transform_shape(img_arr)
img_list.append(img_arr)
im_info = [[img_arr.shape[2], img_arr.shape[3], im_scale]]
self.scale = im_scale
img_list.append(mx.nd.array(im_info))
return img_list