def evaluate():
for image_path in image_paths:
image = cv2.imread(image_path)[:, :, ::-1]
image_fname = osp.split(image_path)[-1]
image_fname_noext = osp.splitext(image_fname)[0]
result_label_fname = 'res_' + image_fname_noext + '.txt'
result_label_path = osp.join(result_label_dir, result_label_fname)
h, w = image.shape[:2]
src_image = image.copy()
image, scale, pad, window = utils.resize_and_pad_image(image, 512)
image = utils.mold_image(image)
image = np.expand_dims(image, axis=0)
# Run object detection
batch_rpn_proposals, batch_rpn_probs = model.keras_model.predict(
[image, anchors], verbose=0)
rpn_proposals = batch_rpn_proposals[0]
rpn_probs = batch_rpn_probs[0]
boxes = utils.denorm_boxes(rpn_proposals, (512, 512))
scores = rpn_probs[..., np.newaxis]
# keep_ix = np.where(rpn_probs > 0.7)[0]
# boxes = boxes[keep_ix]
# scores = rpn_probs[keep_ix]
# for box in boxes:
# cv2.rectangle(src_image, (box[1], box[0]), (box[3], box[2]), (0, 255, 0), 1)
# show_image(src_image, 'image')
# cv2.waitKey(0)
detector = TextDetector()
if inference_mode == 'rpn':
text_boxes = detector.detect(
boxes[:, [1, 0, 3, 2]].astype(np.float32), scores, (512, 512))
# inference_mode == 'text'
else:
text_boxes = detector.detect2(
boxes[:, [1, 0, 3, 2]].astype(np.float32), scores, (512, 512))
text_boxes = text_boxes[:, [0, 1, 4, 5]]
# -left_pad
text_boxes[:, [0, 2]] -= pad[1][0]
# -top_pad
text_boxes[:, [1, 3]] -= pad[0][0]
text_boxes = np.round(text_boxes / scale).astype(np.int32)
text_boxes[:, [0, 2]] = np.clip(text_boxes[:, [0, 2]], 0, w - 1)
text_boxes[:, [1, 3]] = np.clip(text_boxes[:, [1, 3]], 0, h - 1)
with open(result_label_path, 'w') as f:
for text_box in text_boxes:
f.write(','.join(map(str, text_box.tolist())) + '\n')
# image_paths = ['/home/adam/Public/test.jpg']
# image_paths = glob.glob('datasets/art/train_images/*.jpg')
# image_paths = glob.glob('/home/adam/.keras/datasets/text/ctpn/VOCdevkit/VOC2007/JPEGImages/*.jpg')
image_paths = glob.glob('/home/adam/.keras/datasets/icdar2013/focused_scene_text/task12_images/*.jpg')
for image_path in image_paths:
image = cv2.imread(image_path)[:, :, ::-1]
image, scale, pad, window = utils.resize_and_pad_image(image, 512)
src_image = image.copy()
image = utils.mold_image(image)
image = np.expand_dims(image, axis=0)
# Run object detection
start = time.time()
batch_rpn_proposals, batch_rpn_probs = model.keras_model.predict([image, anchors], verbose=0)
rpn_proposals = batch_rpn_proposals[0]
rpn_probs = batch_rpn_probs[0]
boxes = utils.denorm_boxes(rpn_proposals, (512, 512))
scores = rpn_probs[..., np.newaxis]
# keep_ix = np.where(rpn_probs > 0.7)[0]
# boxes = boxes[keep_ix]
# scores = rpn_probs[keep_ix]
# for box in boxes:
# cv2.rectangle(src_image, (box[1], box[0]), (box[3], box[2]), (0, 255, 0), 1)
# show_image(src_image, 'image')
# cv2.waitKey(0)
detector = TextDetector()
if inference_mode == 'rpn':
text_boxes = detector.detect(boxes[:, [1, 0, 3, 2]].astype(np.float32), scores, (512, 512))
# inference_mode == 'text'
else:
text_boxes = detector.detect2(boxes[:, [1, 0, 3, 2]].astype(np.float32), scores, (512, 512))
end = time.time()
def unmold_detections(detections, mrcnn_mask, original_image_shape,
image_shape, window):
"""Reformats the detections of one image from the format of the neural
network output to a format suitable for use in the rest of the
application.
detections: [N, (y1, x1, y2, x2, class_id, score)] in normalized coordinates
mrcnn_mask: [N, height, width, num_classes]
original_image_shape: [H, W, C] Original image shape before resizing
image_shape: [H, W, C] Shape of the image after resizing and padding
window: [y1, x1, y2, x2] Pixel coordinates of box in the image where the real
image is excluding the padding.
Returns:
boxes: [N, (y1, x1, y2, x2)] Bounding boxes in pixels
class_ids: [N] Integer class IDs for each bounding box
scores: [N] Float probability scores of the class_id
masks: [height, width, num_instances] Instance masks
"""
# How many detections do we have?
# Detections array is padded with zeros. Find the first class_id == 0.
zero_ix = np.where(detections[:, 4] == 0)[0]
N = zero_ix[0] if zero_ix.shape[0] > 0 else detections.shape[0]
# Extract boxes, class_ids, scores, and class-specific masks
boxes = detections[:N, :4]
class_ids = detections[:N, 4].astype(np.int32)
scores = detections[:N, 5]
masks = mrcnn_mask[np.arange(N), :, :, class_ids]
# Translate normalized coordinates in the resized image to pixel
# coordinates in the original image before resizing
window = utils.norm_boxes(window, image_shape[:2])
wy1, wx1, wy2, wx2 = window
shift = np.array([wy1, wx1, wy1, wx1])
wh = wy2 - wy1 # window height
ww = wx2 - wx1 # window width
scale = np.array([wh, ww, wh, ww])
# Convert boxes to normalized coordinates on the window
boxes = np.divide(boxes - shift, scale)
# Convert boxes to pixel coordinates on the original image
boxes = utils.denorm_boxes(boxes, original_image_shape[:2])
# Filter out detections with zero area. Happens in early training when
# network weights are still random
exclude_ix = np.where((boxes[:, 2] - boxes[:, 0]) *
(boxes[:, 3] - boxes[:, 1]) <= 0)[0]
if exclude_ix.shape[0] > 0:
boxes = np.delete(boxes, exclude_ix, axis=0)
class_ids = np.delete(class_ids, exclude_ix, axis=0)
scores = np.delete(scores, exclude_ix, axis=0)
masks = np.delete(masks, exclude_ix, axis=0)
N = class_ids.shape[0]
# Resize masks to original image size and set boundary threshold.
full_masks = []
for i in range(N):
# Convert neural network mask to full size mask
full_mask = utils.unmold_mask(masks[i], boxes[i], original_image_shape)
full_masks.append(full_mask)
full_masks = np.stack(full_masks, axis=-1)\
if full_masks else np.empty(masks.shape[1:3] + (0,))
return boxes, class_ids, scores, full_masks