rimosso utilizzo image temporanea

modules/image_classification.py

@@ -4,35 +4,26 @@ import keras.models as models
 from flask import jsonify
 import os
 from modules.utilities.utils import allowed_model
-import uuid
 
 IMAGE_SIZE = (224, 224)
 class_names = ['Tubercolosi', 'No_Tubercolosi', 'Pneumonia', 'No_Pneumonia']
 BASE_PATH = './data/'
 MODEL = BASE_PATH + 'model/'
 
-def image_predict(image_path, model_path):
-    image = cv2.imread(image_path)
-    # image = cv2.cvtColor(image, cv2.COLOR_BGR2RGB)
-    image = cv2.resize(image, IMAGE_SIZE)
+def image_predict(numpy_image, model_path):
+    image = cv2.resize(numpy_image, IMAGE_SIZE)
     images = []
     images.append(image)
     images = np.array(images, dtype='float32')
     model = models.load_model(model_path, compile=False)
     predictions = model.predict(images)    # Vector of probabilities
-    pred_labels = np.argmax(predictions, axis= 1)
-    return predictions, pred_labels
+    #pred_labels = np.argmax(predictions, axis= 1)
+    return predictions#, pred_labels
 
 def image_classification(numpy_image):
     model = MODEL + 'medical-image-classification.h5' 
-    myuuid = uuid.uuid4()
-    filename = 'test_' + str(myuuid) + '.png'
-    cv2.imwrite(filename, numpy_image)
-    try:
-        labels, max_label = image_predict(filename, model)
-        response = {}
-        for i, label in enumerate(labels[0]):
-            response[class_names[i]] = "%.4f" % float(label)
-        return response
-    finally:	
-        os.remove(filename)	
+    labels = image_predict(numpy_image, model)
+    response = {}
+    for i, label in enumerate(labels[0]):
+        response[class_names[i]] = "%.4f" % float(label)
+    return response