use input image!

app.py

@@ -20,7 +20,7 @@ def predict(img: np.ndarray) -> str:
     # input: numpy array of image in RGB (see defaults for https://www.gradio.app/docs/#image)
 
     # Text detection models expect an image in BGR format.
-    image = cv2.cvtColor(im_cv, cv2.COLOR_RGB2BGR)
+    image = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
     # N,C,H,W = batch size, number of channels, height, width.
     N, C, H, W = input_layer_ir.shape
     # Resize the image to meet network expected input sizes.