def vis_all_detection(im_array, detections, imdb_classes=None, thresh=0.):
"""
visualize all detections in one image
:param im_array: [b=1 c h w] in rgb
:param detections: [ numpy.ndarray([[x1 y1 x2 y2 score]]) for j in classes ]
:param imdb_classes: list of names in imdb
:param thresh: threshold for valid detections
:return:
"""
import matplotlib.pyplot as plt
import random
im = image_processing.transform_inverse(im_array, config.PIXEL_MEANS)
plt.imshow(im)
for j in range(1, len(imdb_classes)):
color = (random.random(), random.random(), random.random()) # generate a random color
dets = detections[j]
for i in range(dets.shape[0]):
bbox = dets[i, :4]
score = dets[i, -1]
if score > thresh:
rect = plt.Rectangle((bbox[0], bbox[1]),
bbox[2] - bbox[0],
bbox[3] - bbox[1], fill=False,
edgecolor=color, linewidth=2)
plt.gca().add_patch(rect)
plt.gca().annotate('{} {:.3f}'.format(imdb_classes[j], score),
rect.get_xy(), color='w')
plt.show()
def vis_all_detection(im_array, detections, imdb_classes=None, thresh=0.7):
"""
visualize all detections in one image
:param im_array: [b=1 c h w] in rgb
:param detections: [ numpy.ndarray([[x1 y1 x2 y2 score]]) for j in classes ]
:param imdb_classes: list of names in imdb
:param thresh: threshold for valid detections
:return:
"""
import matplotlib.pyplot as plt
import random
im = image_processing.transform_inverse(im_array, config.PIXEL_MEANS)
plt.imshow(im)
for j in range(1, len(imdb_classes)):
color = (random.random(), random.random(), random.random()) # generate a random color
dets = detections[j]
for i in range(dets.shape[0]):
bbox = dets[i, :4]
score = dets[i, -1]
if score > thresh:
rect = plt.Rectangle((bbox[0], bbox[1]),
bbox[2] - bbox[0],
bbox[3] - bbox[1], fill=False,
edgecolor=color, linewidth=3.5)
plt.gca().add_patch(rect)
plt.gca().text(bbox[0], bbox[1] - 2,
'{:s} {:.3f}'.format(imdb_classes[j], score),
bbox=dict(facecolor=color, alpha=0.5), fontsize=12, color='white')
plt.show()
def save_all_detection(im_array, detections, imdb_classes=None, thresh=0.7):
"""
save all detections in one image with result.png
:param im_array: [b=1 c h w] in rgb
:param detections: [ numpy.ndarray([[x1 y1 x2 y2 score]]) for j in classes ]
:param imdb_classes: list of names in imdb
:param thresh: threshold for valid detections
:return:
"""
import random
im = image_processing.transform_inverse(im_array, config.PIXEL_MEANS)
im = im[:, :, ::-1].copy() # back to b,g,r
for j in range(1, len(imdb_classes)):
color = (255 * random.random(), 255 * random.random(),
255 * random.random()) # generate a random color
dets = detections[j]
for i in range(dets.shape[0]):
bbox = dets[i, :4]
score = dets[i, -1]
if score > thresh:
cv2.rectangle(im, (int(round(bbox[0])), int(round(bbox[1]))),
(int(round(bbox[2])), int(round(bbox[3]))),
color, 2)
cv2.putText(im, '%s' % imdb_classes[j], (bbox[0], bbox[1]),
cv2.FONT_HERSHEY_SIMPLEX, 1.0, color, 2)
cv2.imwrite("result.jpg", im)
def vis_detection(im, dets, thresh=0.):
"""
draw detected bounding boxes
:param im: [b, c, h, w] oin rgb
:param dets: only one class, [N * [4 coordinates score]]
:param thresh: thresh for valid detections
:return:
"""
from rcnn.config import config
from helper.processing.image_processing import transform_inverse
import matplotlib.pyplot as plt
inds = np.where(dets[:, -1] >= thresh)[0]
if len(inds) == 0:
return
inds = np.argsort(dets[:, -1])[::-1]
inds = inds[:20]
class_name = 'obj'
fig, ax = plt.subplots(figsize=(12, 12))
im = transform_inverse(im, config.PIXEL_MEANS)
ax.imshow(im, aspect='equal')
for i in inds:
bbox = dets[i, :4]
score = dets[i, -1]
rect = plt.Rectangle((bbox[0], bbox[1]),
bbox[2] - bbox[0],
bbox[3] - bbox[1],
fill=False,
edgecolor='red',
linewidth=3.5)
ax.add_patch(rect)
ax.text(bbox[0],
bbox[1] - 2,
'{:s} {:3f}'.format(class_name, score),
bbox=dict(facecolor='blue', alpha=0.5),
fontsize=14,
color='white')
ax.set_title('{} detections with p({} | box) >= {:.1f}'.format(
class_name, class_name, thresh),
fontsize=14)
plt.axis('off')
plt.tight_layout()
plt.draw()
plt.show()
def save_all_detection(im_array, detections, imdb_classes=None, thresh=0.7):
"""
save all detections in one image with result.png
:param im_array: [b=1 c h w] in rgb
:param detections: [ numpy.ndarray([[x1 y1 x2 y2 score]]) for j in classes ]
:param imdb_classes: list of names in imdb
:param thresh: threshold for valid detections
:return:
"""
import random
im = image_processing.transform_inverse(im_array, config.PIXEL_MEANS)
im = im[:, :, ::-1].copy() # back to b,g,r
for j in range(1, len(imdb_classes)):
color = (255*random.random(), 255*random.random(), 255*random.random()) # generate a random color
dets = detections[j]
for i in range(dets.shape[0]):
bbox = dets[i, :4]
score = dets[i, -1]
if score > thresh:
cv2.rectangle(im, (int(round(bbox[0])), int(round(bbox[1]))), (int(round(bbox[2])), int(round(bbox[3]))), color, 2)
cv2.imwrite("result.jpg", im)
def vis_detection(im, dets, thresh=0.):
"""
draw detected bounding boxes
:param im: [b, c, h, w] oin rgb
:param dets: only one class, [N * [4 coordinates score]]
:param thresh: thresh for valid detections
:return:
"""
from rcnn.config import config
from helper.processing.image_processing import transform_inverse
import matplotlib.pyplot as plt
inds = np.where(dets[:, -1] >= thresh)[0]
if len(inds) == 0:
return
inds = np.argsort(dets[:, -1])[::-1]
inds = inds[:20]
class_name = 'obj'
fig, ax = plt.subplots(figsize=(12, 12))
im = transform_inverse(im, config.PIXEL_MEANS)
ax.imshow(im, aspect='equal')
for i in inds:
bbox = dets[i, :4]
score = dets[i, -1]
rect = plt.Rectangle((bbox[0], bbox[1]),
bbox[2] - bbox[0],
bbox[3] - bbox[1], fill=False,
edgecolor='red', linewidth=3.5)
ax.add_patch(rect)
ax.text(bbox[0], bbox[1] - 2,
'{:s} {:3f}'.format(class_name, score),
bbox=dict(facecolor='blue', alpha=0.5), fontsize=14, color='white')
ax.set_title('{} detections with p({} | box) >= {:.1f}'.format(class_name, class_name, thresh), fontsize=14)
plt.axis('off')
plt.tight_layout()
plt.draw()
plt.show()
