def detect2(self, net, meta, im, thresh=.5, hier_thresh=.5, nms=.45):
darknet2Montreal = {
'orange': ['orange', 'cup', 'bowl', 'knife', 'fork', 'spoon'],
'cake': ['scrubby'],
'bottle': [
'noodles', 'coke', 'sprite', 'sausages', 'pringles',
'chocolate drink'
],
'book': ['crackers', 'chips'],
'cell phone': ['grape juice', 'orange juice']
}
num = c_int(0)
pnum = pointer(num)
dn.predict_image(net, im)
dets = dn.get_network_boxes(net, im.w, im.h, thresh, hier_thresh, None,
0, pnum)
num = pnum[0]
if (nms): dn.do_nms_obj(dets, num, meta.classes, nms)
res = []
for j in range(num):
for i in range(meta.classes):
if dets[j].prob[i] > 0:
b = dets[j].bbox
darknetName = meta.names[i]
if darknetName in darknet2Montreal.keys():
for montrealName in darknet2Montreal[darknetName]:
res.append((montrealName, dets[j].prob[i],
(b.x, b.y, b.w, b.h)))
else:
res.append((darknetName, dets[j].prob[i], (b.x, b.y,
b.w, b.h)))
res = sorted(res, key=lambda x: -x[1])
return res
def detect(net, meta, im, thresh=.2, hier_thresh=0, nms=.4):
# type: (object, METADATA, IMAGE, float, float, float) -> List[YoloData]
"""
Detect the objects in the given image. free_image function is called inside this function.
Therefore the input darkent image is not usable after calling this function.
:param net:
:param meta:
:param im:
:param thresh:
:param hier_thresh:
:param nms:
:return:
"""
num = c_int(0)
pnum = pointer(num)
predict_image(net, im)
dets = get_network_boxes(net, im.w, im.h, thresh, hier_thresh, None, 0, pnum)
num = pnum[0]
if nms:
do_nms_sort(dets, num, meta.classes, nms)
res = []
for j in range(num):
for i in range(meta.classes):
if dets[j].prob[i] > 0:
b = dets[j].bbox
res.append(YoloData(id=i, name=meta.names[i], bbox=BBox(b.x - b.w/2.0, b.y - b.h/2.0, b.w, b.h, dets[j].prob[i])))
res = sorted(res, key=lambda x: -x.bbox.c)
free_image(im)
free_detections(dets, num)
return res
def detect3(net, meta, image, thresh=.5, hier_thresh=.5, nms=.45):
if isinstance(image, bytes):
# image is a filename
# i.e. image = b'/darknet/data/dog.jpg'
im = load_image(image, 0, 0)
else:
# image is an nparray
# i.e. image = cv2.imread('/darknet/data/dog.jpg')
im, image = array_to_image(image)
dn.rgbgr_image(im)
num = dn.c_int(0)
pnum = dn.pointer(num)
dn.predict_image(net, im)
dets = dn.get_network_boxes(net, im.w, im.h, thresh, hier_thresh, None, 0,
pnum)
num = pnum[0]
if nms: dn.do_nms_obj(dets, num, meta.classes, nms)
res = []
for j in range(num):
a = dets[j].prob[0:meta.classes]
if any(a):
ai = np.array(a).nonzero()[0]
for i in ai:
b = dets[j].bbox
res.append(
(meta.names[i], dets[j].prob[i], (b.x, b.y, b.w, b.h)))
res = sorted(res, key=lambda x: -x[1])
if isinstance(image, bytes): free_image(im)
dn.free_detections(dets, num)
return res
def detect(net, meta, image, thresh=.5, hier_thresh=.5, nms=.45):
# im = load_image(image, 0, 0)
t = time.time()
im = array_to_image(image)
print('array_to_image time: {}').format(time.time() - t)
dn.rgbgr_image(im)
num = dn.c_int(0)
pnum = dn.pointer(num)
dn.predict_image(net, im)
dets = dn.get_network_boxes(net, im.w, im.h, thresh, hier_thresh, None, 0,
pnum)
num = pnum[0]
if (nms): dn.do_nms_obj(dets, num, meta.classes, nms)
t = time.time()
res = []
for j in range(num):
for i in range(meta.classes):
if dets[j].prob[i] > 0:
b = dets[j].bbox
res.append(
(meta.names[i], dets[j].prob[i], (b.x, b.y, b.w, b.h)))
res = sorted(res, key=lambda x: -x[1])
# free_image(im)
# free_detections(dets, num)
return res
def detect(self,
image: str,
thresh: float = .5,
hier_thresh: float = .5,
nms: float = .45) -> list:
if isinstance(image, str):
im = dn.load_image(image.encode(), 0, 0)
elif image is None:
return []
else:
arr = image.transpose(2, 0, 1)
c, h, w = arr.shape
arr = (arr/255.0).flatten()
data = dn.c_array(dn.c_float, arr)
im = dn.IMAGE(w, h, c, data)
num = dn.c_int(0)
pnum = dn.pointer(num)
dn.predict_image(self.net, im)
dets = dn.get_network_boxes(
self.net, im.w, im.h, thresh, hier_thresh, None, 0, pnum)
num = pnum[0]
if (nms):
dn.do_nms_obj(dets, num, self.meta.classes, nms)
res = []
for j in range(num):
if dets[j].prob[PERSON_ID] > 0:
bb = dets[j].bbox
res.append((dets[j].prob[PERSON_ID], BBOX(bb)))
res = sorted(res, key=lambda x: -x[0]) # 0 is prob
# dn.free_image(im) # raise double free error
dn.free_detections(dets, num)
return res
def detect_from_memory(net, meta, im, thresh=.5, hier_thresh=.5, nms=.45, debug=False):
"""
Performs the meat of the detection
"""
num = c_int(0)
if debug:
print("Assigned num")
pnum = pointer(num)
if debug:
print("Assigned pnum")
darknet.predict_image(net, im)
if debug:
print("did prediction")
dets = darknet.get_network_boxes(net, im.w, im.h, thresh,
hier_thresh, None, 0, pnum, 1)
if debug:
print("Got dets")
num = pnum[0]
if debug:
print("got zeroth index of pnum")
if nms:
darknet.do_nms_sort(dets, num, meta.classes, nms)
if debug:
print("did sort")
res = []
if debug:
print("about to range")
for j in range(num):
if debug:
print("Ranging on " + str(j) + " of " + str(num))
if debug:
print("Classes: " + str(meta), meta.classes, meta.names)
for i in range(meta.classes):
if debug:
print("Class-ranging on " + str(i) + " of " +
str(meta.classes) + "= " + str(dets[j].prob[i]))
if dets[j].prob[i] > 0:
b = dets[j].bbox
if darknet.altNames is None:
nameTag = meta.names[i]
else:
nameTag = darknet.altNames[i]
if debug:
print("Got bbox", b)
print(nameTag)
print(dets[j].prob[i])
print((b.x, b.y, b.w, b.h))
res.append((nameTag, dets[j].prob[i], (b.x, b.y, b.w, b.h)))
if debug:
print("did range")
res = sorted(res, key=lambda x: -x[1])
if debug:
print("did sort")
darknet.free_detections(dets, num)
if debug:
print("freed detections")
return res
def detect():
print('Loading image')
# Use tmp.jpg for demo purposes
im = dn.load_image(bytes('tmp.jpg', encoding='utf-8'), 0, 0)
num = dn.c_int(0)
pnum = dn.pointer(num)
print('Predicting image')
dn.predict_image(net, im)
print('Getting boxes')
dets = dn.get_network_boxes(net, im.w, im.h, 0.5, 0.5, None, 1, pnum)
print('Marking boxes')
res = []
classes = 1
for j in range(num.value):
for i in range(classes):
if dets[j].prob[i] > 0.75:
b = dets[j].bbox
res.append((b.x, b.y, b.w, b.h))
dn.free_image(im)
dn.free_detections(dets, num)
print('Saving image')
source_img = Image.open('tmp.jpg').convert("RGB")
size = source_img.size
w = size[0]
h = size[1]
draw = ImageDraw.Draw(source_img)
for b in res:
x1 = (b[0] - b[2] / 2.) * w
x2 = (b[0] + b[2] / 2.) * w
y1 = (b[1] - b[3] / 2.) * h
y2 = (b[1] + b[3] / 2.) * h
draw.rectangle(((x1, y1), (x2, y2)), outline="red")
print(b)
source_img.save('tmp.jpg', "JPEG")
def classify(net, meta, im):
# type: (object, METADATA, IMAGE) -> Tuple[str, float]
out = predict_image(net, im)
res = []
for i in range(meta.classes):
res.append((meta.names[i], out[i]))
res = sorted(res, key=lambda x: -x[1])
return res
def detect2(net, meta, im, thresh=.5, hier_thresh=.5, nms=.45):
num = c_int(0)
pnum = pointer(num)
dn.predict_image(net, im)
dets = dn.get_network_boxes(net, im.w, im.h, thresh, hier_thresh, None, 0,
pnum)
num = pnum[0]
if (nms): dn.do_nms_obj(dets, num, meta.classes, nms)
res = []
for j in range(num):
for i in range(meta.classes):
if dets[j].prob[i] > 0:
b = dets[j].bbox
res.append(
(meta.names[i], dets[j].prob[i], (b.x, b.y, b.w, b.h)))
res = sorted(res, key=lambda x: -x[1])
return res
def detect2(net, meta, image, thresh=.5, hier_thresh=.5, nms=.45):
# num = dn.c_int(0)
# pnum = dn.pointer(num)
# dn.predict_image(net, image)
# dets = dn.get_network_boxes(net, image.w, image.h, thresh, hier_thresh, None, 0, pnum)
# num = pnum[0]
# if (nms): dn.do_nms_obj(dets, num, meta.classes, nms);
# res = []
# for j in range(num):
# for i in range(meta.classes):
# if dets[j].prob[i] > 0:
# b = dets[j].bbox
# res.append((meta.names[i], dets[j].prob[i], (b.x, b.y, b.w, b.h)))
# res = sorted(res, key=lambda x: -x[1])
# dn.free_image(image)
# dn.free_detections(dets, num)
# return res
num = dn.c_int(0)
pnum = dn.pointer(num)
dn.predict_image(net, image)
dets = dn.get_network_boxes(net, image.w, image.h, thresh, hier_thresh,
None, 0, pnum)
num = pnum[0]
if nms: dn.do_nms_obj(dets, num, meta.classes, nms)
res = []
for j in range(num):
a = dets[j].prob[0:meta.classes]
if any(a):
ai = np.array(a).nonzero()[0]
for i in ai:
b = dets[j].bbox
res.append(
(meta.names[i], dets[j].prob[i], (b.x, b.y, b.w, b.h)))
res = sorted(res, key=lambda x: -x[1])
if isinstance(image, bytes):
dn.free_image(image) #这步什么情况下执行？多次无法执行的原因是未对这句做判断
print("free image")
# dn.free_image(image)
dn.free_detections(dets, num)
return res
def detect2(net, meta, image, thresh=.5, hier_thresh=.5, nms=.45, drawBoxes = False, alphabet=None):
num = c_int(0)
pnum = pointer(num)
dn.predict_image(net, im)
dets = dn.get_network_boxes(net, image.w, image.h, thresh, hier_thresh, None, 1, pnum)
num = pnum[0]
if (nms): dn.do_nms_obj(dets, num, meta.classes, nms);
res = []
for j in range(num):
for i in range(meta.classes):
if dets[j].prob[i] > 0:
b = dets[j].bbox
res.append((meta.names[i], dets[j].prob[i], (b.x, b.y, b.w, b.h)))
res = sorted(res, key=lambda x: -x[1])
if drawBoxes == True:
dn.draw_detections(image, dets, num, thresh, meta.names, alphabet, meta.classes)
#dn.free_image(image)
#dn.free_detections(dets, num)
return res
def image_classification(image, network, class_names):
width = darknet.network_width(network)
height = darknet.network_height(network)
image_rgb = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
image_resized = cv2.resize(image_rgb, (width, height), interpolation=cv2.INTER_LINEAR)
darknet_image = darknet.make_image(width, height, 3)
darknet.copy_image_from_bytes(darknet_image, image_resized.tobytes())
detections = darknet.predict_image(network, darknet_image)
predictions = [(name, detections[idx]) for idx, name in enumerate(class_names)]
darknet.free_image(darknet_image)
return sorted(predictions, key=lambda x: -x[1])
def darknet_GPU_predict(img, prob=0.5):
imgW, imgH = SIZE
imgResize, fx, fy, dx, dy = letterbox_image(img, SIZE)
im = array_to_image(imgResize)
out = predict_image(tableNet, im)
values = []
for i in range(2 * imgW * imgH):
values.append(out[i])
out = np.array(values).reshape((2, imgH, imgW))
#out = exp(out)
out = out[:, dy:, dx:]
return out, fx, fy
def detect2(net, meta, image, thresh=.5, hier_thresh=.5, nms=.45):
num = dn.c_int(0)
pnum = dn.pointer(num)
dn.predict_image(net, image)
dets = dn.get_network_boxes(net, image.w, image.h, thresh, hier_thresh,
None, 0, pnum)
num = pnum[0]
if (nms): dn.do_nms_obj(dets, num, meta.classes, nms)
res = []
for j in range(num):
for i in range(meta.classes):
if dets[j].prob[i] > 0:
b = dets[j].bbox
res.append({
"name": meta.names[i].decode("utf-8"),
"conf": dets[j].prob[i],
"box": (b.x, b.y, b.w, b.h)
})
res = sorted(res, key=lambda x: x["name"])
#dn.free_image(im)
dn.free_detections(dets, num)
return res
def detect2(net, meta, cvim, thresh=.7, hier_thresh=.5, nms=.45):
t1 = time.time()
im = nparray_to_image(cvim)
t2 = time.time()
print "cv2 to IMAGE: ", ((t2 - t1) * 1000)
num = c_int(0)
pnum = pointer(num)
dn.predict_image(net, im)
dets = dn.get_network_boxes(net, im.w, im.h, thresh, hier_thresh, None, 0,
pnum)
num = pnum[0]
if (nms): dn.do_nms_obj(dets, num, meta.classes, nms)
res = []
for j in range(num):
for i in range(meta.classes):
if dets[j].prob[i] > 0:
b = dets[j].bbox
res.append(
[meta.names[i], dets[j].prob[i], b.x, b.y, b.w, b.h])
res = sorted(res, key=lambda x: -x[1])
dn.free_image(im)
dn.free_detections(dets, num)
return res
