Face Recognition

1. Experiment Objective

Use MediaPipe Face Detection to detect the positions of faces in the camera image, marking each face region with a rectangular bounding box and displaying the confidence.

2. Experiment Principle

1. Key Terms

2. Technical Architecture

Terminal 1: start agent + camera + joint models
Terminal 2: face recognition

3. Experiment Steps

1. Hardware Preparation

2. Start the agent + camera + joint models (Terminal 1)

xxxxxxxxxx
ros2 launch microros_v2_bringup car_base.launch.py

3. Start face recognition (Terminal 2)

xxxxxxxxxx
ros2 launch microros_v2_mediapipe_pkg face_detection.launch.py

4. Observed Results

• Faces are enclosed in pink rectangular bounding boxes, with the confidence percentage shown above the box
• 6 keypoints (eyes, nose, mouth, ears) are annotated at their corresponding positions

5. Code Walkthrough

Source code paths:

• Launch: ~/microros_ws/src/microros_v2_mediapipe_pkg/launch/face_detection.launch.py
• Node: ~/microros_ws/src/microros_v2_mediapipe_pkg/microros_v2_mediapipe_pkg/FaceDetection.py

1. Core Node Analysis

​x
# Source file: ~/microros_ws/src/microros_v2_mediapipe_pkg/microros_v2_mediapipe_pkg/FaceDetection.py
class FaceDetection(Node):
    def __init__(self):
        super().__init__("face_detection_node")
        self.mp_face_detection = mp.solutions.face_detection
        self.face_detection = self.mp_face_detection.FaceDetection(
            model_selection=0, min_detection_confidence=0.5)
​
    def image_callback(self, msg):
        frame = self.bridge.imgmsg_to_cv2(msg, 'bgr8')
        results = self.face_detection.process(cv.cvtColor(frame, cv.COLOR_BGR2RGB))
        if results.detections:
            for detection in results.detections:
                bbox = detection.location_data.relative_bounding_box
                h, w, _ = frame.shape
                x, y = int(bbox.xmin * w), int(bbox.ymin * h)
                bw, bh = int(bbox.width * w), int(bbox.height * h)
                cv.rectangle(frame, (x, y), (x+bw, y+bh), (255, 0, 255), 2)
                cv.putText(frame, f'{detection.score[0]:.0%}',
                    (x, y-5), cv.FONT_HERSHEY_SIMPLEX, 0.5, (255, 0, 255), 2)
        cv.imshow(self.window_name, frame)

Key logic:

• model_selection=0: short-range model (within 2 meters), suitable for desktop/indoor scenes; =1 is the long-range model
• relative_bounding_box: normalized coordinates (0~1) that must be multiplied by the actual image width/height to convert to pixel coordinates
• The bounding box + confidence percentage are drawn directly on the original frame; the 6 keypoints are obtained via detection.location_data.relative_keypoints

2. Key Parameters

Parameter definition files:

• Launch: ~/microros_ws/src/microros_v2_mediapipe_pkg/launch/face_detection.launch.py
• Node: ~/microros_ws/src/microros_v2_mediapipe_pkg/microros_v2_mediapipe_pkg/FaceDetection.py

3. Node Description

6. Frequently Asked Questions