Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
geoyee d2c3ad6185 | 2 years ago | |
---|---|---|
.. | ||
backbone | 3 years ago | |
layers | 3 years ago | |
README.md | 2 years ago | |
__init__.py | 3 years ago | |
cdminet.py | 2 years ago | |
dsifn.py | 2 years ago | |
dtcdscd.py | 2 years ago | |
fast_scnn.py | 2 years ago | |
snunet_cd.py | 2 years ago | |
stanet.py | 2 years ago | |
unet.py | 2 years ago |
相关模型可以去往ppcd.models下查看,包括参数等等。如果需要使用自建的模型,可在ppcd.models下新建.py文件,完成模型搭建,并在对应的init文件夹下导入自己的模型。有关模型的建立需要注意两点:
模型 | 使用 | 输入图像列表张数 | 返回列表长度 |
---|---|---|---|
Fast-SCNN | ppcd.models.FastSCNN() | $N(N\ge1)$ | 1 / 2 |
UNet | ppcd.models.UNet() | $N(N\ge1)$ | 1 |
SNUNet-CD | ppcd.models.SNUNet() | 2 | 1 / 5 |
DSIFN | ppcd.models.DSIFN() | 2 | 1 |
STANet | ppcd.models.STANet() | 2 | 1 |
*CDMI-Net | ppcd.models.CDMINet() | 2 | 2 |
*DTCDSCD | ppcd.models.CDNet34 | 2 | 3 |
其中上述模型中前5个模型得到的结果均为变化检测图;后2个模型比较特殊,数据组织和训练方式也有所差别,第6个模型以分类的方式进行训练,得到的结果为特征图和分类结果,需要使用阈值等得到变化检测图;第7个模型得到的结果为变化检测图以及两个时段的分割图。
目前的模型与分割模型的定义相似,只是在forward
中需要list作为输入。分割模型的自定义可以参考PaddleSeg自定义模型,在ppcd中构造如下:
import paddle.nn as nn
class NewNet(nn.Layer):
def __init__(self, param1, param2, param3):
pass
def forward(self, imgs):
# 这里的imgs是一个列表,输入图像为[img_1, img_2, img_3, ……]
return [out]
PdRSCD(PaddlePaddle Remote Sensing Change Detection)是一个基于飞桨PaddlePaddle的遥感变化检测的项目,pypi包名为ppcd。目前0.2版本,最新支持图像列表输入的训练和预测,如多期影像、多源影像甚至多期多源影像。可以快速完成分割、变化检测等任务。
Python other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》