MrChen18 / CGSANet

GNU General Public License v3.0
13 stars 1 forks source link

请求 #1

Open YUHsss opened 2 years ago

YUHsss commented 2 years ago

作者您好, 像您的论文中Fig.6这样的只包含边界的预测图是怎样生成的呢?请问方便公布代码吗或者大致说一下流程。 非常感谢!

MrChen18 commented 2 years ago

您好。感谢您对论文的兴趣。Fig.6是论文的边界精度评价图,取示例图像的1/4,也就是256*256大小。先对预测的建筑物区域提取结果和真值做边缘检测生成边界,然后进行精度评价就可以生成这种精度评价图。可以用matlab的imoverlay函数实现图像叠加生成彩色图,也可以利用python写。可视化的效果看图片显示要不要稍微用形态学加粗一下。如果还有疑问,欢迎随时联系~

YUHsss @.***> 于2022年7月22日周五 23:41写道:

作者您好, 像您的论文中Fig.6这样的只包含边界的预测图是怎样生成的呢?请问方便公布代码吗或者大致说一下流程。 非常感谢!

— Reply to this email directly, view it on GitHub https://github.com/MrChen18/CGSANet/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFLFTFAYWZTPWW5OKEZRMKLVVK6LPANCNFSM54L7UMRQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

YUHsss commented 2 years ago

谢谢您的回复---- 回复的原邮件 @.>发送日期2022年08月07日 21:06 @.> @.**@.>主题Re: [MrChen18/CGSANet] 请求 (Issue #1) 您好。感谢您对论文的兴趣。Fig.6是论文的边界精度评价图,取示例图像的1/4,也就是256*256大小。先对预测的建筑物区域提取结果和真值做边缘检测生成边界,然后进行精度评价就可以生成这种精度评价图。可以用matlab的imoverlay函数实现图像叠加生成彩色图,也可以利用python写。可视化的效果看图片显示要不要稍微用形态学加粗一下。如果还有疑问,欢迎随时联系~

YUHsss @.***> 于2022年7月22日周五 23:41写道:

作者您好,

像您的论文中Fig.6这样的只包含边界的预测图是怎样生成的呢?请问方便公布代码吗或者大致说一下流程。

非常感谢!

Reply to this email directly, view it on GitHub

https://github.com/MrChen18/CGSANet/issues/1, or unsubscribe

https://github.com/notifications/unsubscribe-auth/AFLFTFAYWZTPWW5OKEZRMKLVVK6LPANCNFSM54L7UMRQ

.

You are receiving this because you are subscribed to this thread.Message

ID: @.***>

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you authored the thread.Message ID: @.> [ { @.": "http://schema.org", @.": "EmailMessage", "potentialAction": { @.": "ViewAction", "target": "https://github.com/MrChen18/CGSANet/issues/1#issuecomment-1207404712", "url": "https://github.com/MrChen18/CGSANet/issues/1#issuecomment-1207404712", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { @.***": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

YUHsss commented 4 months ago

Love from your friend!