sudipansaha / dcvaVHROptical

Deep Change Vector Analysis
BSD 2-Clause "Simplified" License
52 stars 9 forks source link

test my own picture #15

Open fengfeng-xie opened 6 days ago

fengfeng-xie commented 6 days ago

I want to ask how to use my own pictures(.tif)?This code need mat.But I meet problems when I transform tif into mat

sudipansaha commented 5 days ago

You can easily modify a few lines in the code to suit your requirements. The code essentially processes images in matrix format, so the method of reading the images is flexible with minimal modifications. For instance, take a look at this repository (https://github.com/sudipansaha/unsupervisedPlanetaryCD2021/tree/main), which is based on DCVA. In this case, JPG images are used instead of MAT files.

fengfeng-xie commented 5 days ago

您可以轻松地修改代码中的几行以满足您的要求。该代码实质上以矩阵格式处理图像,因此读取图像的方法非常灵活,只需进行最少的修改。例如,看看这个基于 DCVA 的存储库 (https://github.com/sudipansaha/unsupervisedPlanetaryCD2021/tree/main)。在这种情况下,使用 JPG 图像而不是 MAT 文件。

OK,thank you.My picture is 4 bands.Is it OK?

sudipansaha commented 4 days ago

Use the 4 band model: https://github.com/sudipansaha/dcvaVHROptical/tree/master/trainedNet/RGBIR Your image is acquired by which sensor?

fengfeng-xie commented 4 days ago

使用 4 波段型号:https://github.com/sudipansaha/dcvaVHROptical/tree/master/trainedNet/RGBIR 您的图像由哪个传感器采集?

worldview-2

fengfeng-xie commented 4 days ago

I have modify this code.But the result is wrong. image

sudipansaha commented 4 days ago

You can send me your images to sudipan.saha@scai.iitd.ac.in I will have a look in free time.

fengfeng-xie commented 4 days ago

你可以把你的照片发给我 sudipan.saha@scai.iitd.ac.in 我会在空闲时间看一看。

OK ,thank you