更新灰度图像兼容,修复灰度图像示例工程失败的bug

netrans_py/netrans.py

@@ -133,7 +133,7 @@ class Netrans():
             data (dict): 加载的配置文件内容。
             **kwargs: 需要更新的参数。
         """
-        grey = data['input_meta']['databases'][0]['ports'][0]['preprocess']['preproc_node_params'] == 'IMAGE_GRAY'
+        grey = data['input_meta']['databases'][0]['ports'][0]['preprocess']['preproc_node_params']['preproc_type'] == 'IMAGE_GRAY'
         if 'mean' in kwargs:
             mean = self._format_preprocess_param(kwargs['mean'], grey)
             data = self._upload_config_mean(data, mean)
@@ -141,7 +141,7 @@ class Netrans():
             scale = self._format_preprocess_param(kwargs['scale'], grey)
             data = self._upload_config_scale(data, scale)
         if 'reverse_channel' in kwargs:
-            data = self._upload_config_reverse_channel(data, kwargs['reverse_channel'])
+            data = self._upload_config_reverse_channel(data, kwargs['reverse_channel'],grey)
         return data
 
 
@@ -195,9 +195,17 @@ class Netrans():
         Returns:
             list: 处理后的参数值。
         """
-        if grey:
-            return param
-        return param if isinstance(param, list) else [param] * 3
+        ch = 1 if grey else 3
+        if isinstance(param, (int, float)):
+            return [float(param)] * ch
+        if isinstance(param, (list, tuple)):
+            if len(param) != ch:
+                raise ValueError(
+                    f"灰度图需 1 个值，RGB 图需 3 个值，"
+                    f"当前通道数={ch}，但提供 {len(param)} 个值"
+                )
+            return [float(v) for v in param]
+        raise TypeError("mean / scale 必须是数字或 list/tuple")
 
     def _verify_preprocess_value(self):
         """单元测试中用于判断是否成功修改配置文件中的参数
@@ -241,8 +249,6 @@ class Netrans():
             TypeError: 仅支持量化成 uint8, int8, int16
         """
         if  quantize_type not in ['uint8', 'int8', 'int16']:
-            print(quantize_type)
-            print(type(quantize_type))
             raise TypeError(f"不支持的量化类型: {quantize_type}，仅支持 uint8, int8, int16")
         self.quantize_type = quantize_type
         Quantize(self).quantize_network()