testing trained model with sample image

testing_trained_model.py

@@ -0,0 +1,34 @@
+from keras.models import load_model
+import cv2
+import numpy as np
+
+# Path where the sample image is stored
+FILE_PATH = "testimg.png"
+
+# Path where the trained model is stored
+MODEL_PATH = "rock-paper-scissors-trained.h5"
+
+# Defining class map for each label
+CLASS_MAP = {
+    0: "rock",
+    1: "paper",
+    2: "scissors",
+}
+def mapper(val):
+    return CLASS_MAP[val]
+
+# Loading the trained model
+model = load_model(MODEL_PATH)
+
+# Preparing the image
+img = cv2.imread(FILE_PATH)
+img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
+img = cv2.resize(img, (148, 148))
+
+# Predicting the name
+pred = model.predict(np.array([img]))
+pred_label = np.argmax(pred[0])
+label_value = mapper(pred_label)
+
+# Predicting the final result
+print("Predicted: {}".format(label_value))