我正在使用以下代码解码base64图像：

def string_to_image(base64_string):
    decoded = base64.b64decode(base64_string)
    np_data = np.frombuffer(decoded, np.uint8)
    img = cv2.imdecode(np_data, cv2.IMREAD_UNCHANGED)
    return img

我们的目标是从请求主体接收一个图像，对其进行解码，使用tensorflow调整其大小，使用模型对其进行预测，并返回一个响应，说明该图像是什么：

image_base64 = request.json['image']
decoded_image = string_to_image(image_base64)
image_resized = tf.image.resize(decoded_image, (256, 256))
model = load_model('src/models/mymodel.h5')
result = model.predict(np.expand_dims(image_resized/255, 0))

但是，我得到错误ValueError: Input 0 of layer "sequential_2" is incompatible with the layer: expected shape=(None, 256, 256, 3), found shape=(None, 256, 256, 4)。
我不知道如何将Shape值从“4”更改为“3”。
我尝试了以下方法：

image_resized = tf.image.resize(decoded_image, (256, 256, 3))

但我得到了'size' must be a 1-D Tensor of 2 elements: new_height, new_width。
我也试探着：

image_resized = cv2.resize(decoded_image, (256,256,3))

但是我得到OpenCV(4.6.0) :-1: error: (-5:Bad argument) in function 'resize'
Overload resolution failed:
- Can't parse 'dsize'. Expected sequence length 2, got 3
- Can't parse 'dsize'. Expected sequence length 2, got 3
请帮助：（