yolov5图像归一化

yolov5在图像归一化方面采用的是将图像resize到的范围。具体来说，它将图像的像素值除以255.0，将原本在0-255范围内的像素值转换为在0.0-1.0范围内的值。这个处理操作可以在yolov5的detect.py文件的第67行找到，对图像进行归一化的代码如下：img /= 255.0 # 0 - 255 to 0.0 - 1.0。这样处理后，图像就可以作为网络输入进行后续的处理了。 [1<span class="em">1</span><span class="em">2</span><span class="em">3</span>
#### 引用[.reference_title]
- *1* *2* [Yolov5知识点](https://blog.csdn.net/zicai_jiayou/article/details/109744654)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v92^chatsearchT0_1"}}] [.reference_item style="max-width: 50%"]
- *3* [yolov5数据增强、代码解读](https://blog.csdn.net/weixin_41012399/article/details/119143889)[target="_blank" data-report-click={"spm":"1018.2226.3001.9630","extra":{"utm_source":"vip_chatgpt_common_search_pc_result","utm_medium":"distribute.pc_search_result.none-task-cask-2~all~insert_cask~default-1-null.142^v92^chatsearchT0_1"}}] [.reference_item style="max-width: 50%"]
[ .reference_list ]