diff --git a/solutions/sem02/lesson04/task1.py b/solutions/sem02/lesson04/task1.py
index 1b5526c1..94941f59 100644
--- a/solutions/sem02/lesson04/task1.py
+++ b/solutions/sem02/lesson04/task1.py
@@ -2,16 +2,43 @@
 
 
 def pad_image(image: np.ndarray, pad_size: int) -> np.ndarray:
-    # ваш код
-    return image
-
-
-def blur_image(
-    image: np.ndarray,
-    kernel_size: int,
-) -> np.ndarray:
-    # ваш код
-    return image
+    if pad_size < 1:
+        raise ValueError
+
+    elif image.ndim == 2:
+        h, w = image.shape
+        new_h = h + pad_size * 2
+        new_w = w + pad_size * 2
+
+        padded = np.zeros((new_h, new_w), dtype=image.dtype)
+        padded[pad_size : pad_size + h, pad_size : pad_size + w] = image
+
+    elif image.ndim == 3:
+        h, w, c = image.shape
+        new_h = h + pad_size * 2
+        new_w = w + pad_size * 2
+
+        padded = np.zeros((new_h, new_w, c), dtype=image.dtype)
+        padded[pad_size : pad_size + h, pad_size : pad_size + w, :] = image
+
+    return padded
+
+
+def blur_image(image: np.ndarray, kernel_size: int) -> np.ndarray:
+    if kernel_size % 2 == 0 or kernel_size < 1:
+        raise ValueError
+    if kernel_size == 1:
+        return image
+
+    padding = kernel_size // 2
+    padded = pad_image(image, padding)
+    shifted = [
+        padded[i : i + image.shape[0], j : j + image.shape[1], ...]
+        for i in range(kernel_size)
+        for j in range(kernel_size)
+    ]
+    blurred = np.mean(shifted, axis=0)
+    return blurred.astype(image.dtype)
 
 
 if __name__ == "__main__":
diff --git a/solutions/sem02/lesson04/task2.py b/solutions/sem02/lesson04/task2.py
index be9a2288..8a4c300c 100644
--- a/solutions/sem02/lesson04/task2.py
+++ b/solutions/sem02/lesson04/task2.py
@@ -5,6 +5,25 @@ def get_dominant_color_info(
     image: np.ndarray[np.uint8],
     threshold: int = 5,
 ) -> tuple[np.uint8, float]:
-    # ваш код
+    if threshold < 1:
+        raise ValueError("threshold must be positive")
 
-    return 0, 0
+    pixels = image.flatten()
+    counts = np.zeros(256, dtype=np.int64)
+    for color in range(256):
+        mask = pixels == color
+        counts[color] = np.sum(mask)
+
+    max_pixels = 0
+    best_color = 0
+    for color in range(256):
+        if counts[color] != 0:
+            low = max(0, color - threshold + 1)
+            high = min(255, color + threshold - 1)
+            pixels = np.sum(counts[low : high + 1])
+
+            if pixels > max_pixels:
+                max_pixels = pixels
+                best_color = color
+
+    return np.uint8(best_color), (max_pixels / pixels.size) * 100