multiple faces

app.py

@@ -82,5 +82,39 @@ def recognize_image():
 
     return jsonify({'error': 'Student not recognized'}), 404
 
+@app.route('/rec_multi_face', methods=['POST'])
+def recognize_multiple_faces():
+    if 'face' not in request.files:
+        return jsonify({'error': 'Image not provided'}), 400
+
+    image = request.files['face']
+    # Load the uploaded image file into a numpy array
+    unknown_image = face_recognition.load_image_file(image)
+
+    # Detect face locations in the image
+    face_locations = face_recognition.face_locations(unknown_image)
+    # Get the face encodings for each detected face
+    unknown_encodings = face_recognition.face_encodings(unknown_image, face_locations)
+
+    if len(unknown_encodings) > 0:
+        matches = []
+
+        for unknown_encoding in unknown_encodings:
+            for student_id, known_encoding in students_db.items():
+                # Calculate the distance between the known encoding and the uploaded image encoding
+                distance = face_recognition.face_distance([known_encoding], unknown_encoding)[0]
+                # Convert distance to accuracy score (the closer the distance, the higher the accuracy)
+                accuracy_score = (1 - distance) * 100
+
+                if accuracy_score > 50:  # Threshold to filter out low-confidence matches
+                    matches.append({'student_id': student_id, 'accuracy_score': accuracy_score})
+
+        # Sort matches by accuracy score in descending order
+        matches = sorted(matches, key=lambda x: x['accuracy_score'], reverse=True)
+
+        return jsonify({'number_of_faces': len(unknown_encodings), 'matches': matches}), 200
+
+    return jsonify({'error': 'No recognizable faces found', 'number_of_faces': 0}), 404
+
 if __name__ == '__main__':
     app.run(debug=True, host="0.0.0.0", port=5005)