def initialize_subscr_topics(self):
# Initialize all the subscriber topics
# self.h5_dvc_subscr_obj = ecal.subscriber(json_data['device_request'])
self.h5_chnl_subscr_obj = ecal.subscriber(json_data['channel_request'])
self.h5_img_subscr_obj = ecal.subscriber(json_data['hfl_request'])
self.h5_pixelwriter_obj = ecal.subscriber(
json_data['pixelwrite_request'])
def initialize_subscr_topics(self):
# Initialize all the subscriber topics
self.pnt_cld_dvc_subscr_obj = ecal.subscriber(
json_data['pointcloud_device_request'])
self.pnt_cld_data_subscr_obj = ecal.subscriber(
json_data['pointcloud_data_request'])
self.pnt_cld_end_subscr_obj = ecal.subscriber(
json_data['pointcloud_end_response'])
def __init__(self):
# Subscribe to the eCAL message to get the bsig path and the
# signal names
# self.bsig_path = bsig_pth
#
ecal.initialize(sys.argv, "Python signal value publisher")
# Subscribe to RadarSignalRequest topic
self.bsig_subscr_obj = ecal.subscriber(topic_name="BsigSignalNames")
self.bsig_pub_obj = ecal.publisher(topic_name="BsigSignalValues")
self.bsig_tmstamp_lst = []
self.sig_name_lst = []
self.subscribe_ecal_msgs()
def subscribe_signl_names():
ecal.initialize(sys.argv, "Detector mask subscriber")
subscribe_sig_names_obj = ecal.subscriber(topic_name="Pixel_Response")
rdr_sig_response = AlgoInterface_pb2.LabelResponse()
while ecal.ok():
ret, msg, time = subscribe_sig_names_obj.receive(500)
print("---:: ", ret, msg, time, type(msg))
if msg is not None:
rdr_sig_response.ParseFromString(msg)
object_attibute_lst = rdr_sig_response.NextAttr
# print("object_attibute :: ", object_attibute_lst)
for evry_obj_attr in object_attibute_lst:
track_id = evry_obj_attr.trackID
# print("track_id ::> ", track_id)
class_obj = evry_obj_attr.type.object_class
# print("class_obj :: ", class_obj)
x1 = evry_obj_attr.ROI[0].X
y1 = evry_obj_attr.ROI[0].Y
x2 = evry_obj_attr.ROI[1].X
y2 = evry_obj_attr.ROI[1].Y
x3 = evry_obj_attr.ROI[2].X
y3 = evry_obj_attr.ROI[2].Y
x4 = evry_obj_attr.ROI[3].X
y4 = evry_obj_attr.ROI[3].Y
print("ordinates :: ", x1, x2, x3, x4, y1, y2, y3, y4)
mask = evry_obj_attr.mask
with open('img.txt', 'w+') as fhandle:
fhandle.write(str(mask))
nparr = np.fromstring(mask, np.uint8)
print("nparr :: ", nparr)
re_img_np_ary = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
img_shape = re_img_np_ary.shape
print("shape::", img_shape)
cv2.imwrite('color_img.jpg', re_img_np_ary)
cv2.imshow('Color image', re_img_np_ary)
cv2.waitKey(0)
cv2.destroyAllWindows()
import json
import cv2
if getattr(sys, 'frozen', False):
os.chdir(sys._MEIPASS)
with open('topics.json') as data_file:
json_data = json.load(data_file)
# print(json_data)
request = str(json_data['request'])
response = str(json_data['response'])
h5_filename = str(json_data['h5_filename'])
ecal.initialize(sys.argv, "HFL data publisher")
hfl_subscr_obj = ecal.subscriber(request)
hfl_publs_obj = ecal.publisher(topic_name=response)
hfl_req_proto_obj = imageservice_pb2.ImageRequest()
hfl_resp_proto_obj = imageservice_pb2.HFLResponse()
class H5ReaderSequence(object):
def __init__(self, fname):
self.reset(fname)
self.lastLoadFilename = ""
def reset(self, fname):
if not os.path.isfile(fname):
print("Error - H5 file not available: %s!" % fname)
sys.exit()
def initialize_subscr_topics(self):
# Initialize all the subscriber topics
# self.lt5_img_subscr_obj = ecal.subscriber(self.json_data['image_request'])
self.lt5_img_subscr_obj = ecal.subscriber(self.tracker_request)
self.lt5_finl_subscr_obj = ecal.subscriber(self.json_data['algo_end_response'])
FROZEN_MODEL = "mask_rcnn_coco.h5"
TOPICS_JSON = 'topics.json'
with open(TOPICS_JSON) as data_file:
json_data = json.load(data_file)
# print(json_data)
request = str(json_data['image_request'])
response = str(json_data['image_response'])
vis_flag = True if str(json_data['visualization']) == "True" else False
full_eff_flag = True if str(json_data['full_efficiency']) == "True" else False
ecal.initialize(sys.argv, "Mask RCNN MS COCO detector")
ld_req_obj = imageservice_pb2.ImageResponse()
subscriber_obj = ecal.subscriber(topic_name=request)
lbl_response_obj = AlgoInterface_pb2.LabelResponse()
publisher_roi_obj = ecal.publisher(topic_name=response)
class InferenceConfig(coco.CocoConfig):
# Set batch size to 1 since we'll be running inference on
# one image at a time. Batch size = GPU_COUNT * IMAGES_PER_GPU
GPU_COUNT = 1
IMAGES_PER_GPU = 1
if not full_eff_flag:
IMAGE_MIN_DIM = 512
IMAGE_MAX_DIM = 512