How to crop and save segmented images

[ki580]

Hi, I am trying to get semantic segmentation working in practice and would like to cut out a color coded segmented area and save it individually.
How can I do this after getting the masks information

[cajdotxa71262]

that is what I want to ask

[Bruce337f]

me too,that is what I want to ask!

[ki580]

I can save crop image this code

output_directory = 'your output file'

for i in range(len(mask)):
x, y, width, height = mask[i]['bbox']
cropped_image = image[int(y):int(y+height), int(x):int(x+width)]
filename = os.path.join(output_directory, str(i) + '.png')
cv2.imwrite(filename, cropped_image)

[skycitywing]

我可以保存裁剪图像此代码

output_directory =“您的输出文件”

对于 i in range（len（mask））： x， y， width， height = mask[i]['bbox'] cropped_image = image[int（y）：int（y+height）， int（x）：int（x+width）] 文件名 = os.path.join（output_directory， str（i） + '.png'） cv2.imwrite（文件名， cropped_image）

I hope someone can provide an example on colab.
After getting the mask file and the original image, but don't know how to use the information about the mask boundaries for cropping.

[Github-Yilei]

this might be useful: #221