cheng
300 lines
12 KiB
Python
300 lines
12 KiB
Python
"""
|
||
示例 2：双传感器并行处理（多进程）
|
||
|
||
本示例演示如何使用 multiprocessing 同时运行两个传感器：
|
||
- 每个传感器运行在独立进程中
|
||
- 每个进程拥有独立的 CUDA 上下文
|
||
- 通过 Process 和 Queue 在主进程中汇总结果并显示
|
||
|
||
架构说明：
|
||
进程 1（传感器 0）：采集 → 光流 → HHD → 接触检测 → 队列
|
||
进程 2（传感器 1）：采集 → 光流 → HHD → 接触检测 → 队列
|
||
↓
|
||
主进程： 显示 ← ───────────────┘
|
||
|
||
适用场景：多指触觉传感器、多视角采集，以及需要 GPU 加速的多路处理。
|
||
|
||
注意事项：
|
||
- Windows 下必须在 if __name__ == '__main__' 中调用 mp.freeze_support()
|
||
- 每个传感器进程都会占用独立的 GPU 显存
|
||
|
||
SDK version: 0.3.1
|
||
"""
|
||
import os
|
||
os.environ["OPENCV_VIDEOIO_MSMF_ENABLE_HW_TRANSFORMS"] = "0"
|
||
import cv2
|
||
import orisys
|
||
import numpy as np
|
||
import multiprocessing as mp
|
||
from multiprocessing import Queue, Process
|
||
import time
|
||
import argparse
|
||
|
||
|
||
def parse_video_source(value):
|
||
"""Convert camera index strings like "0" to int; keep paths as strings."""
|
||
if isinstance(value, str):
|
||
value = value.strip()
|
||
if value.isdigit():
|
||
return int(value)
|
||
return value
|
||
|
||
|
||
def sensor_worker(sensor_id, vid_src, config_name, rotation_config_path, result_queue, stop_event, target_fps=None):
|
||
"""
|
||
传感器工作进程。
|
||
|
||
Args:
|
||
sensor_id: 传感器编号
|
||
vid_src: 视频源（文件路径或摄像头索引）
|
||
config_name: 配置名称
|
||
result_queue: 结果队列，用于向主进程发送数据
|
||
stop_event: 停止事件，用于通知工作进程退出
|
||
target_fps: 目标帧率；设置后将对处理循环限速，使多路传感器 FPS 更接近
|
||
"""
|
||
try:
|
||
# 在每个进程中创建独立的传感器实例
|
||
sensor = orisys.Sensor(
|
||
vid_src,
|
||
isstitch=True,
|
||
cuda=True,
|
||
verbose=False, # 减少子进程输出
|
||
config_name=config_name,
|
||
rotation_config_path=rotation_config_path
|
||
)
|
||
|
||
print(
|
||
f"传感器 {sensor_id} 已启动，视频源={vid_src!r}({type(vid_src).__name__})，"
|
||
f"方向配置={rotation_config_path}"
|
||
+ (f"（目标帧率：{target_fps} FPS）" if target_fps else "")
|
||
)
|
||
|
||
frame_interval = 1.0 / target_fps if target_fps and target_fps > 0 else 0.0
|
||
|
||
while not stop_event.is_set():
|
||
t_start = time.perf_counter()
|
||
try:
|
||
# 步骤 1：获取图像
|
||
sensor.get_img()
|
||
|
||
# 步骤 2：计算形变场
|
||
sensor.compute_deformation()
|
||
|
||
# 步骤 3：计算接触区域（如需要可取消注释）
|
||
# sensor.compute_contact()
|
||
|
||
# 步骤 4：读取结果
|
||
fps, flow, vnormal, img, contour, centroid, depth_map = sensor.read_info(
|
||
sensor.info.FPS,
|
||
sensor.info.VRAW,
|
||
sensor.info.VNORMAL,
|
||
sensor.info.IMG,
|
||
sensor.info.CONTOUR,
|
||
sensor.info.CENTROID,
|
||
sensor.info.DEPTH
|
||
)
|
||
|
||
# 步骤 5：整理结果字典（确保数据可序列化）
|
||
data = {
|
||
'sensor_id': sensor_id,
|
||
'fps': fps,
|
||
'flow': flow.copy() if flow is not None else None,
|
||
'vnormal': vnormal.copy() if vnormal is not None else None,
|
||
'img': img.copy() if img is not None else None,
|
||
'contour': contour,
|
||
'centroid': centroid,
|
||
'depth_map': depth_map.copy() if depth_map is not None else None,
|
||
'timestamp': time.time()
|
||
}
|
||
|
||
# 步骤 6：将结果发送到主进程（非阻塞）
|
||
try:
|
||
result_queue.put_nowait(data)
|
||
except Exception:
|
||
# 如果队列已满，则丢弃当前帧，避免阻塞工作进程
|
||
pass
|
||
|
||
except Exception as e:
|
||
print(f"传感器 {sensor_id} 处理失败：{e}")
|
||
time.sleep(0.1)
|
||
|
||
# 步骤 7：按 target_fps 限速，使多路传感器 FPS 更接近
|
||
if frame_interval > 0:
|
||
elapsed = time.perf_counter() - t_start
|
||
sleep_time = frame_interval - elapsed
|
||
if sleep_time > 0.001: # 至少 1ms 才 sleep，避免无意义调度
|
||
time.sleep(sleep_time)
|
||
|
||
# 释放资源
|
||
sensor.disconnect()
|
||
print(f"传感器 {sensor_id} 已停止。")
|
||
|
||
except Exception as e:
|
||
print(f"传感器 {sensor_id} 初始化失败：{e}")
|
||
result_queue.put({'sensor_id': sensor_id, 'error': str(e)})
|
||
|
||
|
||
def main():
|
||
# =========================================================================
|
||
# 1. 配置多路传感器
|
||
# =========================================================================
|
||
# 目标帧率：设为正数时，两路传感器会按该帧率限速；设为 None 则不限速
|
||
|
||
parser = argparse.ArgumentParser()
|
||
parser.add_argument("--video1","-v1", type=str, default="0", help="视频源：摄像头编号或视频文件路径")
|
||
parser.add_argument("--video2","-v2", type=str, default="0", help="视频源：摄像头编号或视频文件路径")
|
||
parser.add_argument("--rotation1", "-r1", type=str, default="./config/rotation_config_0.json", help="传感器 0 的方向配置")
|
||
parser.add_argument("--rotation2", "-r2", type=str, default="./config/rotation_config_1.json", help="传感器 1 的方向配置")
|
||
args = parser.parse_args()
|
||
|
||
target_fps = 30
|
||
|
||
num_sensors = 2
|
||
sensor_configs = [
|
||
{
|
||
'sensor_id': 0,
|
||
'vid_src': parse_video_source(args.video1),
|
||
'config_name': './config/ddjx01.json',
|
||
'rotation_config_path': args.rotation1
|
||
},
|
||
{
|
||
'sensor_id': 1,
|
||
'vid_src': parse_video_source(args.video2), # 可替换为其他视频源或摄像头索引
|
||
'config_name': './config/ddjx01.json',
|
||
'rotation_config_path': args.rotation2
|
||
},
|
||
]
|
||
|
||
# =========================================================================
|
||
# 2. 创建进程间通信对象
|
||
# =========================================================================
|
||
# 适当增大队列容量，以缓冲更多帧并减轻主进程偶发延迟带来的丢帧
|
||
result_queues = [Queue(maxsize=30) for _ in sensor_configs]
|
||
stop_events = [mp.Event() for _ in sensor_configs]
|
||
|
||
# =========================================================================
|
||
# 3. 启动工作进程
|
||
# =========================================================================
|
||
processes = []
|
||
for i, config in enumerate(sensor_configs):
|
||
p = Process(
|
||
target=sensor_worker,
|
||
args=(i, config['vid_src'], config['config_name'], config['rotation_config_path'],
|
||
result_queues[i], stop_events[i], target_fps)
|
||
)
|
||
p.start()
|
||
processes.append(p)
|
||
|
||
print(f"\n已启动 {len(processes)} 个传感器进程。")
|
||
print("按键说明：'q' 退出程序。")
|
||
|
||
# 保存每个传感器的最新一帧数据
|
||
latest_data = [None] * len(sensor_configs)
|
||
|
||
# 控制打印频率，减少 I/O 开销
|
||
print_counter = 0
|
||
print_interval = 30 # 每 30 帧打印一次
|
||
|
||
# 控制按键检查频率
|
||
key_check_counter = 0
|
||
key_check_interval = 5 # 每 5 帧检查一次按键
|
||
|
||
try:
|
||
# =====================================================================
|
||
# 4. 主循环：从队列获取结果并显示
|
||
# =====================================================================
|
||
while True:
|
||
# 从所有队列中快速获取最新数据，只保留最后一帧
|
||
for i, queue in enumerate(result_queues):
|
||
# 清空队列，仅保留最后一帧
|
||
data = None
|
||
# 读取所有可用数据，只保留最新结果
|
||
while True:
|
||
try:
|
||
data = queue.get_nowait()
|
||
except:
|
||
break
|
||
|
||
if data is not None:
|
||
latest_data[i] = data
|
||
|
||
# 显示所有传感器的最新结果
|
||
for data in latest_data:
|
||
if data is None or 'error' in data:
|
||
continue
|
||
|
||
sid = data['sensor_id']
|
||
|
||
# 显示形变矢量场
|
||
if data['img'] is not None and data['flow'] is not None:
|
||
arrows = orisys.util.draw_arrows(
|
||
data['img'],
|
||
data['flow'],
|
||
threshold=5,
|
||
grid_spacing=10,
|
||
arrow_scale=1.0
|
||
)
|
||
cv2.imshow(f"传感器 {sid} - 形变矢量场".encode("gbk"), arrows)
|
||
|
||
# # 显示触碰点和区域
|
||
# if data['img'] is not None:
|
||
# image_with_foe = orisys.util.draw_contact(
|
||
# data['img'],
|
||
# data['contour'],
|
||
# data['centroid']
|
||
# )
|
||
# cv2.imshow(f"Contact Centroid {sid}", image_with_foe)
|
||
|
||
# # 显示深度图
|
||
# if data['depth_map'] is not None and data['depth_map'].size > 0:
|
||
# div_abs = data['depth_map']
|
||
# if div_abs.max() > div_abs.min():
|
||
# div_normalized = ((div_abs - div_abs.min()) /
|
||
# (div_abs.max() - div_abs.min()) * 255).astype(np.uint8)
|
||
# else:
|
||
# div_normalized = np.zeros_like(div_abs, dtype=np.uint8)
|
||
# cv2.imshow(f"Depth Map {sid}",
|
||
# cv2.applyColorMap(div_normalized, cv2.COLORMAP_JET))
|
||
|
||
# 降低打印频率，减少主循环中的 I/O 开销
|
||
print_counter += 1
|
||
if print_counter >= print_interval:
|
||
for data in latest_data:
|
||
if data is not None and 'error' not in data:
|
||
sid = data['sensor_id']
|
||
print(f"传感器 {sid}：FPS={data['fps']:5.1f}")
|
||
print_counter = 0
|
||
|
||
# 降低按键检查频率，减少额外开销
|
||
key_check_counter += 1
|
||
if key_check_counter >= key_check_interval:
|
||
key = cv2.waitKey(1) & 0xFF
|
||
if key == ord('q'):
|
||
print("\n正在退出示例程序...")
|
||
break
|
||
key_check_counter = 0
|
||
|
||
finally:
|
||
# =====================================================================
|
||
# 5. 释放资源
|
||
# =====================================================================
|
||
# 通知所有工作进程停止
|
||
for event in stop_events:
|
||
event.set()
|
||
|
||
# 等待所有工作进程结束
|
||
for p in processes:
|
||
p.join(timeout=2)
|
||
if p.is_alive():
|
||
p.terminate()
|
||
p.join()
|
||
|
||
cv2.destroyAllWindows()
|
||
print("所有传感器进程已停止。")
|
||
|
||
|
||
if __name__ == '__main__':
|
||
# Windows 下运行多进程示例所需的初始化
|
||
mp.freeze_support()
|
||
main()
|