found perfect parameters 2/2

inference_test.py

@@ -98,12 +98,31 @@ class TextProcessor:
         return final_text
 
 
-def adjust_gamma(image, gamma=1.0):
+import cv2
+import numpy as np
+
+def adjust_gamma(image, gamma=1.0, alpha=1.0, beta=0):
+    """
+    Adjust the gamma and contrast of a grayscale image.
+
+    Parameters:
+        image (numpy.ndarray): Input grayscale image.
+        gamma (float): Gamma correction factor. Values < 1 darken the image, > 1 brighten it.
+        alpha (float): Contrast control. 1.0 means no change, < 1.0 reduces contrast, > 1.0 increases contrast.
+        beta (int): Brightness control. Positive values brighten the image, negative values darken it.
+
+    Returns:
+        numpy.ndarray: Adjusted image.
+    """
+    # Apply gamma correction
     invGamma = 1.0 / gamma
     table = np.array([((i / 255.0) ** invGamma) * 255 for i in np.arange(0, 256)]).astype("uint8")
-    return cv2.LUT(image, table)
-    
+    gamma_corrected = cv2.LUT(image, table)
 
+    # Adjust contrast and brightness
+    adjusted = cv2.convertScaleAbs(gamma_corrected, alpha=alpha, beta=beta)
+
+    return adjusted
 
 
 def preprocess_image(image_path):
@@ -111,7 +130,7 @@ def preprocess_image(image_path):
     gray = cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)
 
     # Convert grayscale to 3-channel RGB by duplicating the gray channel
-    img = adjust_gamma(cv2.cvtColor(gray, cv2.COLOR_GRAY2RGB),gamma=2)
+    img = adjust_gamma(cv2.cvtColor(gray, cv2.COLOR_GRAY2RGB),gamma=2,alpha=1.5,beta=-0.5) # perfect tunning
     mat_in = ncnn.Mat.from_pixels_resize(
         img, ncnn.Mat.PixelType.PIXEL_RGB, img.shape[1], img.shape[0], 224, 224
     )