gray-scale shift

inference_test.py

@@ -102,13 +102,17 @@ class TextProcessor:
 
 
 def preprocess_image(image_path):
-    img = cv2.imread(image_path)
+    # Load image in grayscale (1-channel)
+    gray = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
+
+    # Convert grayscale to 3-channel RGB by duplicating the gray channel
+    img = cv2.cvtColor(gray, cv2.COLOR_GRAY2RGB)
     mat_in = ncnn.Mat.from_pixels_resize(
         img, ncnn.Mat.PixelType.PIXEL_RGB, img.shape[1], img.shape[0], 224, 224
     )
     
-    mean_vals = [0.5, 0.5, 0.5] * 3
-    norm_vals = [1/255, 1/255, 1/255] * 3
+    mean_vals = [0.5, 0.5, 0.5] * 1
+    norm_vals = [1/255, 1/255, 1/255] * 1
     mat_in.substract_mean_normalize(mean_vals, norm_vals)
     return mat_in
 
@@ -127,11 +131,11 @@ def run_inference(mat_in, model_path="model"):
     try:
         ret, out0 = ex.extract("out0")
         print("Output shape:", out0.shape)
-        predictions = np.array(out0)
+        predictions = np.squeeze(np.array(out0))
       #  ids = np.argmax(predictions, axis=-1)
         ids = argmax_last_axis(predictions) 
         print("argmax'ed: ", ids)
-        return np.squeeze(ids)
+        return ids
     finally:
         del ex, net