use bicubic

inference_test.py

@@ -107,13 +107,11 @@ class TextProcessor:
         return final_text
 
 
-def bilinear(img):
+def bicubic(img):
     h, w = img.shape[:2]
-    # Create identity maps
-    map_x, map_y = np.meshgrid(np.arange(w, dtype=np.float32),
-                               np.arange(h, dtype=np.float32))
-    # Remap with INTER_LINEAR
-    return cv2.remap(img, map_x, map_y, interpolation=cv2.INTER_LINEAR)
+    M = np.array([[1, 0, 0],
+                  [0, 1, 0]], dtype=np.float32)
+    return cv2.warpAffine(img, M, (w, h), flags=cv2.INTER_CUBIC)
 
 
 def adjust_gamma(image, gamma=1.0, alpha=1.0, beta=0):
@@ -143,7 +141,7 @@ def adjust_gamma(image, gamma=1.0, alpha=1.0, beta=0):
 def preprocess_image(image_path):
     # Load image in grayscale (1-channel)
     gray = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
-    resampled = bilinear(gray) 
+    resampled = bicubic(gray) 
     # Convert grayscale to 3-channel RGB by duplicating the gray channel
     img = adjust_gamma(cv2.cvtColor(resampled, cv2.COLOR_GRAY2RGB),gamma=1.3,alpha=1.65,beta=-2) # perfect tunning
     mat_in = ncnn.Mat.from_pixels_resize(