search

abzargar / COVID-Classifier

An efficient machine learning model to assist in the diagnosis of COVID-19 infection in chest x-ray images
11 stars 13 forks source link

mistake in your code #2

Open kiranrao2019 opened 3 years ago

kiranrao2019 commented 3 years ago

Hi @abzargar ,

you did an exceptional work. I have 2 question from you regarding your code. In your process_image.py file you did preprocessing for all of your images and then saved the preprocessing images in to the computer. But in your extract_feature.py file, in the for loop you are again doing image rescaling. I think you are doing it by mistake at this point because if you are using the preprocessing images for extracting features then you already did this rescaling in the process_image.py so why need to do the rescaling again? I think you are doing it by mistake please correct me if I am wrong. Or, on the side I am wondering may be in extract_feature.py you are not considering the preprocessing images file and considering the original images.

Waiting for your response. Thanks Kiran

abzargar commented 3 years ago

Thank you for your message and your attention to my paper code. You are correct. I am preprocessing images and then using them in the feature extraction process but I did not make any mistake in the extracting feature algorithm. There, I just rescale pixel values (which are in the range between 0 and 255, UINT8) to the range between 0 and1 (float64) which is necessary for the next steps of the process like the compute_14_features function.

Sincerely,

Abolfazl Zargari

Ph.D. student and research assistant

Electrical and Computer Engineering Department

The University of California, Santa Cruz Email: abzargar@ucsc.edu

On Tue, Nov 3, 2020 at 2:50 PM kiranrao2019 notifications@github.com wrote:

Hi there,

you did an exceptional work. I have 2 question from you regarding your code. In your process_image.py file you did preprocessing for all of your images and then saved the preprocessing images in to the computer. But in your extract_feature.py file, in the for loop you are again doing rescaling. I think you are doing it by mistake at this point because if you are using the preprocessing images for extracting features then you already did this rescaling in the process_image.py so why need to do the rescaling again? I think you are doing it by mistake please correct me if I am wrong. Or, on the side I am wondering may be in extract_feature.py you are not considering the preprocessing images file and considering the original images.

Waiting for your response. Thanks Kiran

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://mailtrack.io/trace/link/5c422746c5af32cac0e6fb54a1207866821c0c2f?url=https%3A%2F%2Fgithub.com%2Fabzargar%2FCOVID-Classifier%2Fissues%2F2&userId=2833009&signature=e16f3057c02c083e, or unsubscribe https://mailtrack.io/trace/link/aa41ae22e46604ca4e012065eda35261dd47e9a3?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FANZFO64RQUZXWVWHGJYB4NDSOCCK7ANCNFSM4TJKNS6A&userId=2833009&signature=826a00842a81f52a .

asmajavaidd commented 3 years ago

Hi @abzargar,

Thank you for taking the time to respond to my above query. I have one more question from you why did not you do the segmentation on your images after the preprocessing step? Is it because lung X-ray images do not require segmentation because the covid virus effect all part of your lungs.?

I really appreciate your time.