Merge pull request #2124 from theshubh007/master

Face_recognition_api_Added

Author: Yashbhadiyadra

Flask_Apis/Image_recognition_from_File_format.py

@@ -7,11 +7,13 @@
 
 app = Flask(__name__)
 
+face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + "haarcascade_frontalface_default.xml")
+
 @app.route('/')
 def home():
     return jsonify({'message':'Welcome to Flask Apis'})
 
-@app.route('/img',methods=['POST'])
+@app.route('/image_Compare',methods=['POST'])
 def predict():
     file1 = request.files['file1']
     file2 = request.files['file2']
@@ -37,6 +39,47 @@ def predict():
     # Return the similarity percentage in a JSON response
     return jsonify({'similarity_percentage': similarity_percentage})    
 
+@app.route('/face_recognize',methods=['POST'])
+def predictface():
+     # Get the uploaded files from the request
+    file1 = request.files['file1']
+    file2 = request.files['file2']
+
+   # Read the images using OpenCV directly from the request files
+    img1 = cv2.imdecode(np.frombuffer(file1.read(), np.uint8), cv2.IMREAD_COLOR)
+    img2 = cv2.imdecode(np.frombuffer(file2.read(), np.uint8), cv2.IMREAD_COLOR)
+
+    # Convert the images to grayscale
+    gray_img1 = cv2.cvtColor(img1, cv2.COLOR_BGR2GRAY)
+    gray_img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2GRAY)
+
+    # Detect faces in the images
+    faces1 = face_cascade.detectMultiScale(gray_img1, scaleFactor=1.1, minNeighbors=5)
+    faces2 = face_cascade.detectMultiScale(gray_img2, scaleFactor=1.1, minNeighbors=5)
+
+    # Compare only the first detected face in each image
+    if len(faces1) > 0 and len(faces2) > 0:
+        x1, y1, w1, h1 = faces1[0]
+        x2, y2, w2, h2 = faces2[0]
+
+        # Extract the face regions from the images
+        face1 = gray_img1[y1:y1+h1, x1:x1+w1]
+        face2 = gray_img2[y2:y2+h2, x2:x2+w2]
+
+        # Resize the face regions to the same dimensions
+        resized_face1 = cv2.resize(face1, (face2.shape[1], face2.shape[0]))
+
+        # Calculate the structural similarity index between the face regions
+        score = ssim(resized_face1, face2, full=True)[0]
+
+        # Convert the similarity score to a percentage
+        similarity_percentage = score * 100
+
+        # Return the similarity percentage in a JSON response
+        return jsonify({'similarity_percentage': similarity_percentage})
+
+    else:
+        return jsonify({'similarity_percentage': 'Could not detect faces in both images.'})
 
 if __name__ == '__main__':
     app.run(debug=True)