how to create light backgroup for an image?

[cgz004]

hello, I read your paper Preprocessing of Breast Cancer Images to Create Datasets for Deep-CNN. There is a sentence in the paper The light background is applied using the OpenCV (Open Computer Vision) python program. I tried several ways to create light backgroup, but I can't do it. Can you show me the code about create light backgroup? thanks.

[algalib007]

hello, I read your paper Preprocessing of Breast Cancer Images to Create Datasets for Deep-CNN. There is a sentence in the paper The light background is applied using the OpenCV (Open Computer Vision) python program. I tried several ways to create light backgroup, but I can't do it. Can you show me the code about create light backgroup? thanks.

here you go 👇 from cv2_rolling_ball import subtract_background_rolling_ball

radius=5 final_img, lightbackground = subtract_background_rolling_ball(ims, radius, light_background = True, use_paraboloid = False, do_presmooth =False)

cv2.imshow("lightbackground ", lightbackground )

[ABRB13]

Hello Al,

Hope you are doing well.

Sorry for the late reply.

I cannot share the full code, as it is part of a grant. I suggest you use 'ImageJ' software to apply the rolling ball algorithm and others for the light background (it works well).

Regards, Abhijith

On Mon, Apr 4, 2022 at 9:51 AM Al Galib @.***> wrote:

hello, I read your paper Preprocessing of Breast Cancer Images to Create Datasets for Deep-CNN. There is a sentence in the paper The light background is applied using the OpenCV (Open Computer Vision) python program. I tried several ways to create light backgroup, but I can't do it. Can you show me the code about create light backgroup? thanks.

here you go 👇 from cv2_rolling_ball import subtract_background_rolling_ball

final_img, lightbackground = subtract_background_rolling_ball(ims, radius, light_background = True, use_paraboloid = False, do_presmooth =False)

cv2.imshow("lightbackground ", lightbackground )

— Reply to this email directly, view it on GitHub https://github.com/ABRB13/BreastCancer/issues/1#issuecomment-1086984751, or unsubscribe https://github.com/notifications/unsubscribe-auth/AO2QME7BA72HMXSVMJYGRPTVDIYWTANCNFSM5FL3BRRQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

[scher711]

In your paper, for the rolling ball algoritgm you used a ball size of 5 for 1024x1024 image. Would you recommend a smaller ball size for low resolution images? Is this algorithm even applicable if the resolution is much lower?