From 257f37aaefc31dc26d7724e6f87b0169a623a66a Mon Sep 17 00:00:00 2001
From: Javi Ribera <jprat@purdue.edu>
Date: Sun, 16 Sep 2018 19:39:03 -0400
Subject: [PATCH] encapsulate thresholding into a function

Former-commit-id: d78ecf6b20b15e9e622ce4b1cb08bf151b8c0483
---
 object-locator/locate.py | 13 +------------
 object-locator/train.py  | 12 +-----------
 object-locator/utils.py  | 30 ++++++++++++++++++++++++++++++
 3 files changed, 32 insertions(+), 23 deletions(-)

diff --git a/object-locator/locate.py b/object-locator/locate.py
index 75c4770..5f5c15c 100644
--- a/object-locator/locate.py
+++ b/object-locator/locate.py
@@ -216,18 +216,7 @@
     
     # The estimated map must be thresholded to obtain estimated points
     for tau, df_out in zip(args.taus, df_outs):
-        if tau == -1:
-            # Otsu thresholding
-            minn, maxx = est_map_numpy_origsize.min(), est_map_numpy_origsize.max()
-            est_map_origsize_scaled = ((est_map_numpy_origsize - minn)/(maxx - minn)*255) \
-                .round().astype(np.uint8).squeeze()
-            tau_otsu, mask = cv2.threshold(est_map_origsize_scaled,
-                                           0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
-            tau_otsu = minn + (tau_otsu/255)*(maxx - minn)
-            # print(f'Otsu selected tau={tau_otsu}')
-        else:
-            # Thresholding with a fixed threshold tau
-            mask = cv2.inRange(est_map_numpy_origsize, tau, 1)
+        mask, _ = utils.threshold(est_map_numpy_origsize, tau)
         coord = np.where(mask > 0)
         y = coord[0].reshape((-1, 1))
         x = coord[1].reshape((-1, 1))
diff --git a/object-locator/train.py b/object-locator/train.py
index ce92dbc..5b4f765 100644
--- a/object-locator/train.py
+++ b/object-locator/train.py
@@ -328,17 +328,7 @@
 
         # The estimated map must be thresholed to obtain estimated points
         # Otsu thresholding
-        est_map_numpy = est_maps[0, :, :].to(device_cpu).numpy()
-        est_map_numpy_origsize = \
-            skimage.transform.resize(est_map_numpy,
-                                     output_shape=origsize,
-                                     mode='constant')
-        minn, maxx = est_map_numpy_origsize.min(), est_map_numpy_origsize.max()
-        est_map_origsize_scaled = ((est_map_numpy_origsize - minn)/(maxx - minn)*255) \
-            .round().astype(np.uint8).squeeze()
-        tau_otsu, mask = cv2.threshold(est_map_origsize_scaled,
-                                       0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
-        tau_otsu = minn + (tau_otsu/255)*(maxx - minn)
+        mask, _ = utils.threshold(est_map_numpy_origsize, tau=-1)
 
         # Validation metrics
         coord = np.where(mask > 0)
diff --git a/object-locator/utils.py b/object-locator/utils.py
index 397f39f..a12ccf4 100644
--- a/object-locator/utils.py
+++ b/object-locator/utils.py
@@ -1,5 +1,6 @@
 import torch
 import numpy as np
+import cv2
 
 class Normalizer():
     def __init__(self, new_size_height, new_size_width):
@@ -36,3 +37,32 @@ def unnormalize(self, coordinates_yx_normalized, orig_img_size):
 
         return coordinates_yx_unnormalized
 
+def threshold(array, tau):
+    """
+    Threshold an array using either hard thresholding or Otsu thresholding.
+
+    :param array: Array to threshold.
+    :param tau: (float) Threshold to use.
+                Values above tau become 1, and values below tau become 0.
+                If -1, use Otsu thresholding.
+    :return: Tuple, where first element is the binary mask, and the second one
+             is the threshold used. When using Otsu thresholding, this threshold will be
+             is obtained adaptively according to the values of the input array.
+             
+    """
+    if tau == -1:
+        # Otsu thresholding
+        minn, maxx = array.min(), array.max()
+        array_scaled = ((array - minn)/(maxx - minn)*255) \
+            .round().astype(np.uint8).squeeze()
+        tau, mask = cv2.threshold(array_scaled,
+                                  0, 255, cv2.THRESH_BINARY + cv2.THRESH_OTSU)
+        tau = minn + (tau/255)*(maxx - minn)
+        # print(f'Otsu selected tau={tau_otsu}')
+    else:
+        # Thresholding with a fixed threshold tau
+        mask = cv2.inRange(array, tau, 1)
+
+    return mask, tau
+
+