def __init__(self, params, calib_data, hourly_data, daily_data, monthly_data): self.logger = logging.getLogger('pywws.Tasks.RegularTasks') self.params = params self.calib_data = calib_data self.hourly_data = hourly_data self.daily_data = daily_data self.monthly_data = monthly_data # get directories self.work_dir = self.params.get('paths', 'work', '/tmp/weather') self.template_dir = self.params.get( 'paths', 'templates', os.path.expanduser('~/weather/templates/')) self.graph_template_dir = self.params.get( 'paths', 'graph_templates', os.path.expanduser('~/weather/graph_templates/')) # create calibration object self.calibrator = Calib(self.params) # create templater object self.templater = Template.Template(self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data) # create plotter objects self.plotter = Plot.GraphPlotter(self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) self.roseplotter = WindRose.RosePlotter(self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) # directory of service uploaders self.services = dict() # create a YoWindow object self.yowindow = YoWindow.YoWindow(self.calib_data) # get local time's offset from UTC, without DST now = self.calib_data.before(datetime.max) if not now: now = datetime.now() # was kenmc utcnow time_offset = Local.utcoffset(now) - Local.dst(now) print "Time Offset = %s" % time_offset # get daytime end hour, in UTC self.day_end_hour = eval(params.get('config', 'day end hour', '21')) self.day_end_hour = (self.day_end_hour - (time_offset.seconds / 3600)) % 24 # convert config from underground/metoffice to new services for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): services = eval(self.params.get(section, 'services', '[]')) for svc in ('underground', 'metoffice'): if self.params.get(section, svc) == 'True': if svc not in services: services.append(svc) self.params._config.remove_option(section, svc) self.params.set(section, 'services', str(services)) # create service uploader objects for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): for service in eval(self.params.get(section, 'services', '[]')): if service not in self.services: self.services[service] = ToService(self.params, self.calib_data, service_name=service)
def __init__(self, params, calib_data, hourly_data, daily_data, monthly_data): self.logger = logging.getLogger('pywws.Tasks.RegularTasks') self.params = params self.calib_data = calib_data self.hourly_data = hourly_data self.daily_data = daily_data self.monthly_data = monthly_data # get directories self.work_dir = self.params.get('paths', 'work', '/tmp/weather') self.template_dir = self.params.get( 'paths', 'templates', os.path.expanduser('~/weather/templates/')) self.graph_template_dir = self.params.get( 'paths', 'graph_templates', os.path.expanduser('~/weather/graph_templates/')) # create calibration object self.calibrator = Calib(self.params) # create templater object self.templater = Template.Template( self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data) # create plotter objects self.plotter = Plot.GraphPlotter( self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) self.roseplotter = WindRose.RosePlotter( self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) # directory of service uploaders self.services = dict() # create a YoWindow object self.yowindow = YoWindow.YoWindow(self.calib_data) # get local time's offset from UTC, without DST now = self.calib_data.before(datetime.max) if not now: now = datetime.now() # was kenmc utcnow time_offset = Local.utcoffset(now) - Local.dst(now) print "Time Offset = %s" % time_offset # get daytime end hour, in UTC self.day_end_hour = eval(params.get('config', 'day end hour', '21')) self.day_end_hour = (self.day_end_hour - (time_offset.seconds / 3600)) % 24 # convert config from underground/metoffice to new services for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): services = eval(self.params.get(section, 'services', '[]')) for svc in ('underground', 'metoffice'): if self.params.get(section, svc) == 'True': if svc not in services: services.append(svc) self.params._config.remove_option(section, svc) self.params.set(section, 'services', str(services)) # create service uploader objects for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): for service in eval(self.params.get(section, 'services', '[]')): if service not in self.services: self.services[service] = ToService( self.params, self.calib_data, service_name=service)
def calibrate_data(logger, params, raw_data, calib_data): """'Calibrate' raw data, using a user-supplied function.""" start = calib_data.before(datetime.max) if start is None: start = datetime.min start = raw_data.after(start + SECOND) if start is None: return start del calib_data[start:] calib_data.flush() calibrator = Calib(params) day = None for data in raw_data[start:]: idx = data['idx'] if idx.day != day: logger.info("calib: %s", idx.isoformat(' ')) day = idx.day calib_data[idx] = calibrator.calib(data) calib_data.flush() return start
def _rgb_hand_seg(self, hand, box, asd, depth): box = Calib().box_depth2rgb(box, depth) roi = self._crop_box(hand, box) roi = cv2.cvtColor(roi, cv2.COLOR_BGR2YCR_CB) roi = cv2.threshold(roi[:, :, 2], 0, 255, cv2.THRESH_OTSU)[1] kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE, (3, 3)) asd = cv2.dilate(asd, kernel, iterations=3) asd = cv2.resize(asd, (roi.shape[1], roi.shape[0]), interpolation=cv2.cv.CV_INTER_LINEAR) roi[asd == 0] = 0 roi = cv2.medianBlur(roi, 5) return roi
class RegularTasks(object): def __init__(self, params, calib_data, hourly_data, daily_data, monthly_data): self.logger = logging.getLogger('pywws.Tasks.RegularTasks') self.params = params self.calib_data = calib_data self.hourly_data = hourly_data self.daily_data = daily_data self.monthly_data = monthly_data # get directories self.work_dir = self.params.get('paths', 'work', '/tmp/weather') self.template_dir = self.params.get( 'paths', 'templates', os.path.expanduser('~/weather/templates/')) self.graph_template_dir = self.params.get( 'paths', 'graph_templates', os.path.expanduser('~/weather/graph_templates/')) # create calibration object self.calibrator = Calib(self.params) # create templater object self.templater = Template.Template(self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data) # create plotter objects self.plotter = Plot.GraphPlotter(self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) self.roseplotter = WindRose.RosePlotter(self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) # directory of service uploaders self.services = dict() # create a YoWindow object self.yowindow = YoWindow.YoWindow(self.calib_data) # get local time's offset from UTC, without DST now = self.calib_data.before(datetime.max) if not now: now = datetime.now() # was kenmc utcnow time_offset = Local.utcoffset(now) - Local.dst(now) print "Time Offset = %s" % time_offset # get daytime end hour, in UTC self.day_end_hour = eval(params.get('config', 'day end hour', '21')) self.day_end_hour = (self.day_end_hour - (time_offset.seconds / 3600)) % 24 # convert config from underground/metoffice to new services for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): services = eval(self.params.get(section, 'services', '[]')) for svc in ('underground', 'metoffice'): if self.params.get(section, svc) == 'True': if svc not in services: services.append(svc) self.params._config.remove_option(section, svc) self.params.set(section, 'services', str(services)) # create service uploader objects for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): for service in eval(self.params.get(section, 'services', '[]')): if service not in self.services: self.services[service] = ToService(self.params, self.calib_data, service_name=service) def has_live_tasks(self): yowindow_file = self.params.get('live', 'yowindow', '') if yowindow_file: return True for template in eval(self.params.get('live', 'twitter', '[]')): return True for service in eval(self.params.get('live', 'services', '[]')): return True for template in eval(self.params.get('live', 'plot', '[]')): return True for template in eval(self.params.get('live', 'text', '[]')): return True return False def do_live(self, data): data = self.calibrator.calib(data) OK = True yowindow_file = self.params.get('live', 'yowindow', '') if yowindow_file: self.yowindow.write_file(yowindow_file, data) for template in eval(self.params.get('live', 'twitter', '[]')): if not self.do_twitter(template, data): OK = False for service in eval(self.params.get('live', 'services', '[]')): self.services[service].RapidFire(data, True) uploads = [] for template in eval(self.params.get('live', 'plot', '[]')): upload = self.do_plot(template) if upload and upload not in uploads: uploads.append(upload) for template in eval(self.params.get('live', 'text', '[]')): upload = self.do_template(template, data) if upload not in uploads: uploads.append(upload) if uploads: if not Upload.Upload(self.params, uploads): OK = False for file in uploads: os.unlink(file) return OK def do_tasks(self): sections = ['logged'] now = self.calib_data.before(datetime.max) if not now: now = datetime.now() threshold = now.replace(minute=(now.minute / 15) * 15, second=0, microsecond=0) last_update = self.params.get_datetime('hourly', 'last update') if (not last_update) or (last_update < threshold): # time to do hourly tasks sections.append('hourly') # set 12 hourly threshold threshold -= timedelta(hours=(threshold.hour - self.day_end_hour) % 12) last_update = self.params.get_datetime('12 hourly', 'last update') if (not last_update) or (last_update < threshold): # time to do 12 hourly tasks sections.append('12 hourly') # set daily threshold threshold -= timedelta(hours=(threshold.hour - self.day_end_hour) % 24) last_update = self.params.get_datetime('daily', 'last update') if (not last_update) or (last_update < threshold): # time to do daily tasks sections.append('daily') OK = True for section in sections: for template in eval(self.params.get(section, 'twitter', '[]')): if not self.do_twitter(template): OK = False for section in sections: yowindow_file = self.params.get(section, 'yowindow', '') if yowindow_file: self.yowindow.write_file(yowindow_file) break all_services = list() for section in sections: for service in eval(self.params.get(section, 'services', '[]')): if service not in all_services: all_services.append(service) for service in all_services: self.services[service].Upload(True) uploads = [] for section in sections: for template in eval(self.params.get(section, 'plot', '[]')): upload = self.do_plot(template) if upload and upload not in uploads: uploads.append(upload) for template in eval(self.params.get(section, 'text', '[]')): upload = self.do_template(template) if upload not in uploads: uploads.append(upload) if uploads: if not Upload.Upload(self.params, uploads): OK = False for file in uploads: os.unlink(file) if OK: for section in sections: self.params.set(section, 'last update', now.isoformat(' ')) return OK def do_twitter(self, template, data=None): import ToTwitter twitter = ToTwitter.ToTwitter(self.params) self.logger.info("Templating %s", template) input_file = os.path.join(self.template_dir, template) tweet = self.templater.make_text(input_file, live_data=data) self.logger.info("Tweeting") return twitter.Upload(tweet[:140]) def do_plot(self, template): self.logger.info("Graphing %s", template) input_file = os.path.join(self.graph_template_dir, template) output_file = os.path.join(self.work_dir, os.path.splitext(template)[0]) if self.plotter.DoPlot(input_file, output_file) == 0: return output_file elif self.roseplotter.DoPlot(input_file, output_file) == 0: return output_file return None def do_template(self, template, data=None): self.logger.info("Templating %s", template) input_file = os.path.join(self.template_dir, template) output_file = os.path.join(self.work_dir, template) self.templater.make_file(input_file, output_file, live_data=data) return output_file
class RegularTasks(object): def __init__(self, params, calib_data, hourly_data, daily_data, monthly_data): self.logger = logging.getLogger('pywws.Tasks.RegularTasks') self.params = params self.calib_data = calib_data self.hourly_data = hourly_data self.daily_data = daily_data self.monthly_data = monthly_data # get directories self.work_dir = self.params.get('paths', 'work', '/tmp/weather') self.template_dir = self.params.get( 'paths', 'templates', os.path.expanduser('~/weather/templates/')) self.graph_template_dir = self.params.get( 'paths', 'graph_templates', os.path.expanduser('~/weather/graph_templates/')) # create calibration object self.calibrator = Calib(self.params) # create templater object self.templater = Template.Template( self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data) # create plotter objects self.plotter = Plot.GraphPlotter( self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) self.roseplotter = WindRose.RosePlotter( self.params, self.calib_data, self.hourly_data, self.daily_data, self.monthly_data, self.work_dir) # directory of service uploaders self.services = dict() # create a YoWindow object self.yowindow = YoWindow.YoWindow(self.calib_data) # get local time's offset from UTC, without DST now = self.calib_data.before(datetime.max) if not now: now = datetime.now() # was kenmc utcnow time_offset = Local.utcoffset(now) - Local.dst(now) print "Time Offset = %s" % time_offset # get daytime end hour, in UTC self.day_end_hour = eval(params.get('config', 'day end hour', '21')) self.day_end_hour = (self.day_end_hour - (time_offset.seconds / 3600)) % 24 # convert config from underground/metoffice to new services for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): services = eval(self.params.get(section, 'services', '[]')) for svc in ('underground', 'metoffice'): if self.params.get(section, svc) == 'True': if svc not in services: services.append(svc) self.params._config.remove_option(section, svc) self.params.set(section, 'services', str(services)) # create service uploader objects for section in ('live', 'logged', 'hourly', '12 hourly', 'daily'): for service in eval(self.params.get(section, 'services', '[]')): if service not in self.services: self.services[service] = ToService( self.params, self.calib_data, service_name=service) def has_live_tasks(self): yowindow_file = self.params.get('live', 'yowindow', '') if yowindow_file: return True for template in eval(self.params.get('live', 'twitter', '[]')): return True for service in eval(self.params.get('live', 'services', '[]')): return True for template in eval(self.params.get('live', 'plot', '[]')): return True for template in eval(self.params.get('live', 'text', '[]')): return True return False def do_live(self, data): data = self.calibrator.calib(data) OK = True yowindow_file = self.params.get('live', 'yowindow', '') if yowindow_file: self.yowindow.write_file(yowindow_file, data) for template in eval(self.params.get('live', 'twitter', '[]')): if not self.do_twitter(template, data): OK = False for service in eval(self.params.get('live', 'services', '[]')): self.services[service].RapidFire(data, True) uploads = [] for template in eval(self.params.get('live', 'plot', '[]')): upload = self.do_plot(template) if upload and upload not in uploads: uploads.append(upload) for template in eval(self.params.get('live', 'text', '[]')): upload = self.do_template(template, data) if upload not in uploads: uploads.append(upload) if uploads: if not Upload.Upload(self.params, uploads): OK = False for file in uploads: os.unlink(file) return OK def do_tasks(self): sections = ['logged'] now = self.calib_data.before(datetime.max) if not now: now = datetime.now() threshold = now.replace(minute=(now.minute/15)*15, second=0, microsecond=0) last_update = self.params.get_datetime('hourly', 'last update') if (not last_update) or (last_update < threshold): # time to do hourly tasks sections.append('hourly') # set 12 hourly threshold threshold -= timedelta(hours=(threshold.hour - self.day_end_hour) % 12) last_update = self.params.get_datetime('12 hourly', 'last update') if (not last_update) or (last_update < threshold): # time to do 12 hourly tasks sections.append('12 hourly') # set daily threshold threshold -= timedelta(hours=(threshold.hour - self.day_end_hour) % 24) last_update = self.params.get_datetime('daily', 'last update') if (not last_update) or (last_update < threshold): # time to do daily tasks sections.append('daily') OK = True for section in sections: for template in eval(self.params.get(section, 'twitter', '[]')): if not self.do_twitter(template): OK = False for section in sections: yowindow_file = self.params.get(section, 'yowindow', '') if yowindow_file: self.yowindow.write_file(yowindow_file) break all_services = list() for section in sections: for service in eval(self.params.get(section, 'services', '[]')): if service not in all_services: all_services.append(service) for service in all_services: self.services[service].Upload(True) uploads = [] for section in sections: for template in eval(self.params.get(section, 'plot', '[]')): upload = self.do_plot(template) if upload and upload not in uploads: uploads.append(upload) for template in eval(self.params.get(section, 'text', '[]')): upload = self.do_template(template) if upload not in uploads: uploads.append(upload) if uploads: if not Upload.Upload(self.params, uploads): OK = False for file in uploads: os.unlink(file) if OK: for section in sections: self.params.set(section, 'last update', now.isoformat(' ')) return OK def do_twitter(self, template, data=None): import ToTwitter twitter = ToTwitter.ToTwitter(self.params) self.logger.info("Templating %s", template) input_file = os.path.join(self.template_dir, template) tweet = self.templater.make_text(input_file, live_data=data) self.logger.info("Tweeting") return twitter.Upload(tweet[:140]) def do_plot(self, template): self.logger.info("Graphing %s", template) input_file = os.path.join(self.graph_template_dir, template) output_file = os.path.join(self.work_dir, os.path.splitext(template)[0]) if self.plotter.DoPlot(input_file, output_file) == 0: return output_file elif self.roseplotter.DoPlot(input_file, output_file) == 0: return output_file return None def do_template(self, template, data=None): self.logger.info("Templating %s", template) input_file = os.path.join(self.template_dir, template) output_file = os.path.join(self.work_dir, template) self.templater.make_file(input_file, output_file, live_data=data) return output_file
total_time_cost = time.time() - time_start_all print(pointXYZ_raw.shape, camera_xyz.shape, camera_uv.shape) print("transform time cost:", transform_time_cost) print("display time cost:", display_time_cost) print("total time cost: ", total_time_cost) print() if __name__ == '__main__': try: rospy.init_node("fusion") image_topic_name = rospy.get_param("~image_topic") lidar_topic_name = rospy.get_param("~lidar_topic") calib = Calib() file_path = rospy.get_param("~calibration_file_path") the_view_number = rospy.get_param("~the_view_number") the_field_of_view = rospy.get_param("~the_field_of_view") the_sensor_height = rospy.get_param("~the_sensor_height") the_view_higher_limit = rospy.get_param("~the_view_higher_limit") the_view_lower_limit = rospy.get_param("~the_view_lower_limit") the_min_distance = rospy.get_param("~the_min_distance") the_max_distance = rospy.get_param("~the_max_distance") jet_color = rospy.get_param("~jet_color") calib.loadcalib(file_path) LidarImage()
def main(data_filename): # Root directory of the project # data_filename = FLAGS.image_filename ROOT_DIR = os.path.dirname(os.path.realpath(__file__)) print(ROOT_DIR) # Directory to save logs and trained model MODEL_DIR = os.path.join(ROOT_DIR, "logs") # Directory to get binary and image data DATA_DIR = os.path.join("/".join(ROOT_DIR.split("/")[:-1]), "classify/data") # Local path to trained weights file COCO_MODEL_PATH = os.path.join(ROOT_DIR, "mask_rcnn_coco.h5") # Directory of images to run detection on #IMAGE_DIR = os.path.join(ROOT_DIR, "images") 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 config = InferenceConfig() # config.display() # Create model object in inference mode. model = modellib.MaskRCNN(mode="inference", model_dir=MODEL_DIR, config=config) # Load weights trained on MS-COCO model.load_weights(COCO_MODEL_PATH, by_name=True) # COCO Class names # Index of the class in the list is its ID. For example, to get ID of # the teddy bear class, use: class_names.index('teddy bear') class_names = [ 'BG', 'person', 'bicycle', 'car', 'motorcycle', 'airplane', 'bus', 'train', 'truck', 'boat', 'traffic light', 'fire hydrant', 'stop sign', 'parking meter', 'bench', 'bird', 'cat', 'dog', 'horse', 'sheep', 'cow', 'elephant', 'bear', 'zebra', 'giraffe', 'backpack', 'umbrella', 'handbag', 'tie', 'suitcase', 'frisbee', 'skis', 'snowboard', 'sports ball', 'kite', 'baseball bat', 'baseball glove', 'skateboard', 'surfboard', 'tennis racket', 'bottle', 'wine glass', 'cup', 'fork', 'knife', 'spoon', 'bowl', 'banana', 'apple', 'sandwich', 'orange', 'broccoli', 'carrot', 'hot dog', 'pizza', 'donut', 'cake', 'chair', 'couch', 'potted plant', 'bed', 'dining table', 'toilet', 'tv', 'laptop', 'mouse', 'remote', 'keyboard', 'cell phone', 'microwave', 'oven', 'toaster', 'sink', 'refrigerator', 'book', 'clock', 'vase', 'scissors', 'teddy bear', 'hair drier', 'toothbrush' ] # Load a random image from the images folder #file_names = next(os.walk(IMAGE_DIR))[2] # filename = chooseFile.openFile() filename = os.path.join(DATA_DIR, os.path.join("image", data_filename + ".png")) image = Image.open(os.path.join(filename)) start = time() img = skimage.io.imread(os.path.join(filename)) #image = skimage.io.imread(os.path.join(IMAGE_DIR,file_names)) # Run detection results = model.detect([img], verbose=1) print(type(img)) # Visualize results r = results[0] print("identify image time: " + str(time() - start)) contour = visualize.display_instances(img, r['rois'], r['masks'], r['class_ids'], class_names, r['scores']) print(r['class_ids']) # visualize.display_instances(img, r['rois'], r['masks'], r['class_ids'], class_names, r['scores']) calib = Calib( '/Users/berniewang/annotator/lidarAnnotator/app/classify/calib') im = PIL.Image.open(os.path.join(filename)) w, h = im.size bin_name = os.path.join(DATA_DIR, os.path.join("bin_data", data_filename + ".bin")) scan = np.fromfile(os.path.join(bin_name), dtype=np.float32).reshape( (-1, 4)) print("scan: ", scan[0]) im_coord = calib.velo2img(scan[:, :3], 2).astype(np.int) im_coord2 = [ im_coord[i] for i in range(len(im_coord)) if im_coord[i][0] > 0 and im_coord[i][0] <= w and im_coord[i][1] > 0 and im_coord[i][1] <= h and scan[i][0] >= 0 ] #scan2 = [scan[i] for i in range(len(im_coord)) if im_coord[i][0] > 0 and im_coord[i][0] <= w and im_coord[i][1] > 0 and im_coord[i][1] <= h] scan2 = np.array([0], dtype=np.float32) visible_indices = np.array([0], dtype=np.int) for i in range(len(im_coord)): if im_coord[i][0] > 0 and im_coord[i][0] <= w and im_coord[i][ 1] > 0 and im_coord[i][1] <= h: if scan[i][0] >= 0: scan2 = np.append(scan2, scan[i]) visible_indices = np.append(visible_indices, i) scan2 = np.delete(scan2, 0) visible_indices = np.delete(visible_indices, 0) scan2 = scan2.reshape((int(len(scan2) / 4), 4)) #scan2.tofile("scan2.bin") ''' def ray_tracing_method(x,y,poly): n = len(poly) inside = False p1x,p1y = poly[0] for i in range(n+1): p2x,p2y = poly[i % n] if y > min(p1y,p2y): if y <= max(p1y,p2y): if x <= max(p1x,p2x): if p1y != p2y: xints = (y-p1y)*(p2x-p1x)/(p2y-p1y)+p1x if p1x == p2x or x <= xints: inside = not inside p1x,p1y = p2x,p2y return inside ''' start_time = time() print(len(contour)) bounded_indices = np.array([0], dtype=np.int) for i in range(len(contour)): #print(len(contour)) for q in contour[i][0]: temp = q[0] q[0] = q[1] q[1] = temp #print(contour[i][0]) #if len(contour[i][0])>=130: polygon = [q for q in contour[i][0]] path = mpltPath.Path(polygon) inside2 = path.contains_points(im_coord2) #inside1 = [ray_tracing_method(p[1], p[0], polygon) for p in im_coord2] im_coord3 = [] scan3 = np.array([0], dtype=np.float32) for j in range(len(inside2)): if inside2[j] == True: scan3 = np.append(scan3, scan2[j]) bounded_indices = np.append(bounded_indices, visible_indices[j]) scan3 = np.delete(scan3, 0) print("bounded_indices: ", bounded_indices.shape, bounded_indices) scan3 = scan3.reshape((int(len(scan3) / 4), 4)) print(len(scan3)) print(scan3) class_id = r['class_ids'][i] label = class_names[class_id] if len(scan3) != 0: pass #print(len(scan3)) # commented out # scan3.tofile("sample%s%s.bin"%(label,i)) #scan4 = np.append(scan3,scan) #scan4.tofile("a.bin") class_id = r['class_ids'][i] label = class_names[class_id] bounded_indices = np.delete(bounded_indices, 0) bounded_indices.tofile(os.path.join(ROOT_DIR, "output/indices.bin")) print("Matplotlib contains_points Elapsed time: " + str(time() - start_time)) #getContour() print('OK')
def generate_2d_lidar(): calib = Calib( '/Users/berniewang/annotator/lidarAnnotator/app/classify/calib') # load image img_name = getCoord.getPictureName() # print("image path again: ", os.path.join(image_ab_path+img_name)) im = Image.open(os.path.join(image_ab_path + img_name)) w, h = im.size #im = np.array(im) # load lidar points bin_name = getCoord.getFileName() # print(bin_name) scan = np.fromfile(os.path.join(bin_ab_path + bin_name), dtype=np.float32).reshape((-1, 4)) #im_coord0 = calib.velo2img(scan[:, :3], 2).astype(np.int) x = getCoord.getX() y = getCoord.getY() width = getCoord.getWidth() length = getCoord.getLength() angle = getCoord.getAngle() image_path = [] for i in range(len(x)): a = x[i] b = y[i] c = width[i] / 2 d = length[i] / 2 e = angle[i] out_of_fov = False a_array = np.array([0], dtype=np.float32) for j in range(len(scan)): if scan[j][0] > a - c and scan[j][0] < a + c and scan[j][ 1] > b - d and scan[j][1] < b + d: #if scan[j][0] > a - c * abs(math.cos(e)) and scan[j][0] < a + c * abs(math.cos(e)) and scan[j][1] > b - d * abs(math.cos(e)) and scan[j][1]< b + d * abs(math.cos(e)): a_array = np.append(a_array, scan[j]) #if scan[j][0] > a - c and scan[j][0] < a + c and scan[j][1] > b - d and scan[j][1]< b + d: a_array = np.delete(a_array, 0) a_array = a_array.reshape((int(len(a_array) / 4), 4)) a_array.tofile(os.path.join(current_dir_path, "a.bin")) scan2 = np.fromfile(os.path.join(current_dir_path, 'a.bin'), dtype=np.float32).reshape((-1, 4)) im_coord = calib.velo2img(scan2[:, :3], 2).astype(np.int) # plt.imshow(im) im_coord2 = [ im_coord[i] for i in range(len(im_coord)) if im_coord[i][0] > 0 and im_coord[i][0] <= w and im_coord[i][1] > 0 and im_coord[i][1] <= h ] x_arr = [] y_arr = [] # print("length of im_coord2: ", len(im_coord2)) if len(im_coord2) != 0: x_max = im_coord2[0][0] for i in range(len(im_coord2)): if im_coord2[i][0] > x_max: x_max = im_coord2[i][0] x_min = im_coord2[0][0] for i in range(len(im_coord2)): if im_coord2[i][0] < x_min: x_min = im_coord2[i][0] y_max = im_coord2[0][1] for i in range(len(im_coord2)): if im_coord2[i][1] > y_max: y_max = im_coord2[i][1] y_min = im_coord2[0][1] for i in range(len(im_coord2)): if im_coord2[i][1] < y_min: y_min = im_coord2[i][1] #print(x_max,y_max,x_min,y_min) else: x_max = w y_max = h x_min = 0 y_min = 0 print("your frame can not show on the picture") out_of_fov = True # filename = 'write_data.txt' # with open(filename,'a') as f: # f.write("your frame can not show on the picture\n") box = (x_min, y_min, x_max, y_max) image1 = im.crop(box) #图像裁剪 # image1.show() image1.save( "/Users/berniewang/annotator/lidarAnnotator/app/classify/inception/%d.jpg" % (i + 1)) if not out_of_fov: image_path.append( "/Users/berniewang/annotator/lidarAnnotator/app/classify/inception/%d.jpg" % (i + 1)) else: image_path.append("") for i in range(len(im_coord)): x_arr.append(im_coord[i][0]) y_arr.append(im_coord[i][1]) # plt.scatter(x_arr,y_arr,s = 1) # plt.show() # plt.imshow(np.asarray(image1)) # plt.show() return image_path
def generate_2d_lidar(): calib = Calib(CALIBRATION_PATH) # load image img_name = getCoord.getPictureName() drivename, fname = img_name.split("/") fname = fname.split(".")[0] print("img_name: ", img_name) im = Image.open( os.path.join(DATA_DIR, drivename, IMAGE_PATH, fname) + '.png') w, h = im.size #im = np.array(im) # load lidar points bin_name = getCoord.getFileName() # print(bin_name) scan = np.fromfile(os.path.join(DATA_DIR, drivename, BIN_PATH, fname) + '.bin', dtype=np.float32).astype(float).reshape((-1, 4)) #im_coord0 = calib.velo2img(scan[:, :3], 2).astype(np.int) x = getCoord.getX() y = getCoord.getY() width = getCoord.getWidth() length = getCoord.getLength() angle = getCoord.getAngle() image_path = [] for i in range(len(x)): a = x[i] b = y[i] c = width[i] / 2 d = length[i] / 2 e = angle[i] transformed_pc = (scan[:, :2] - np.array([a, b])).dot( rotation_matrix(-e).T) out_of_fov = False a_array = np.array([0], dtype=np.float32) print(scan, len(scan)) for j in range(len(scan)): # if abs(scan[j][0]) <= c and abs(scan[j][1]) <= d: if abs(transformed_pc[j][0]) <= c and abs( transformed_pc[j][1]) <= d: # if scan[j][0] > a - c * abs(math.cos(e)) and scan[j][0] < a + c * abs(math.cos(e)) and scan[j][1] > b - d * abs(math.cos(e)) and scan[j][1]< b + d * abs(math.cos(e)): a_array = np.append(a_array, scan[j]) #if scan[j][0] > a - c and scan[j][0] < a + c and scan[j][1] > b - d and scan[j][1]< b + d: a_array = np.delete(a_array, 0) a_array = a_array.reshape((int(len(a_array) / 4), 4)) print(a_array, len(a_array)) a_array.astype(np.float32).tofile(os.path.join(CUR_DIR, "a.bin")) scan2 = np.fromfile(os.path.join(CUR_DIR, 'a.bin'), dtype=np.float32).astype(float).reshape((-1, 4)) im_coord = calib.velo2img(scan2[:, :3], 2).astype(np.int) # plt.imshow(im) im_coord2 = [ im_coord[i] for i in range(len(im_coord)) if im_coord[i][0] > 0 and im_coord[i][0] <= w and im_coord[i][1] > 0 and im_coord[i][1] <= h ] x_arr = [] y_arr = [] # print("length of im_coord2: ", len(im_coord2)) if len(im_coord2) != 0: x_max = im_coord2[0][0] for i in range(len(im_coord2)): if im_coord2[i][0] > x_max: x_max = im_coord2[i][0] x_min = im_coord2[0][0] for i in range(len(im_coord2)): if im_coord2[i][0] < x_min: x_min = im_coord2[i][0] y_max = im_coord2[0][1] for i in range(len(im_coord2)): if im_coord2[i][1] > y_max: y_max = im_coord2[i][1] y_min = im_coord2[0][1] for i in range(len(im_coord2)): if im_coord2[i][1] < y_min: y_min = im_coord2[i][1] #print(x_max,y_max,x_min,y_min) else: x_max = w y_max = h x_min = 0 y_min = 0 print("your frame can not show on the picture") out_of_fov = True # filename = 'write_data.txt' # with open(filename,'a') as f: # f.write("your frame can not show on the picture\n") box = (x_min, y_min, x_max, y_max) image1 = im.crop(box) #图像裁剪 # image1.show() image1.save("{}/{}.jpg".format(IMAGE_OUTPUT_PATH, i + 1)) if not out_of_fov: image_path.append("{}/{}.jpg".format(IMAGE_OUTPUT_PATH, i + 1)) else: image_path.append("") for i in range(len(im_coord)): x_arr.append(im_coord[i][0]) y_arr.append(im_coord[i][1]) # plt.imshow(im) # plt.scatter(x_arr,y_arr,s = 1) # plt.show() # plt.imshow(np.asarray(image1)) # plt.show() if not out_of_fov: DIR_PATH = os.path.dirname(os.path.realpath(__file__)) print(type(image1)) io.imsave( os.path.join(DIR_PATH, "../static/images/cropped_image.jpg"), np.array(image1)) return image_path
def loadJson(): image_ab_path = chooseFile.openFile() while True: num_file = getNumFiles(path) for i in getCurrentDirs(): if getNumFiles(path) > num_file and os.path.splitext(i)[1] == ".json": jsonFile = i f = io.open("/Users/bc/Downloads/"+ jsonFile, encoding='utf-8') #设置以utf-8解码模式读取文件,encoding参数必须设置,否则默认以gbk模式读取文件,当文件中包含中文时,会报错 setting = json.load(f) file = setting['frames'][0].get('filename') x = [] y = [] width = [] length = [] family = setting['frames'][0] for j in range(len(family.get('bounding_boxes'))): x.append(family.get('bounding_boxes')[j].get('center').get('x')) y.append(family.get('bounding_boxes')[j].get('center').get('y')) width.append(family.get('bounding_boxes')[j].get('width')) length.append(family.get('bounding_boxes')[j].get('length')) current_dir_path = os.path.dirname(os.path.realpath(__file__)) #image_ab_path = chooseFile.openFile() bin_ab_path = current_dir_path+'/' calib = Calib('/Users/bc/3d-2d/calib') im = Image.open(os.path.join(image_ab_path)) w, h = im.size scan = np.fromfile( os.path.join(bin_ab_path+file), dtype=np.float32).reshape((-1, 4)) a_array = np.array([0],dtype = np.float32) for i in range(len(x)): a = x[i] b = y[i] c = width[i]/2 d = length[i]/2 #e = angle[i] for j in range(len(scan)): if scan[j][0] > a - c and scan[j][0] < a + c and scan[j][1] > b - d and scan[j][1]< b + d: a_array = np.append(a_array,scan[j]) a_array = np.delete(a_array,0) a_array = a_array.reshape((int(len(a_array)/4),4)) im_coord = calib.velo2img(a_array[:, :3], 2).astype(np.int) plt.imshow(im) x_arr = [] y_arr = [] for i in range(len(im_coord)): x_arr.append(im_coord[i][0]) y_arr.append(im_coord[i][1]) plt.scatter(x_arr,y_arr,s = 1) plt.show() num_file = getNumFiles(path)