Open AMSSN opened 2 months ago
I didn’t notice it before, but you’re right. Thank you for your correction.
img_temp = img_current # 1.if you use this
img_temp = copy.deepcopy(img_current) # 3.so we use the deepcopy
img_temp = np.uint8(img_temp) img_temp[img_temp != label_id] = 0 # 2.there will cover the img_temp, and cover the img_current, in next loop, the data will be changed in "mask2txt_custom.py" line 31-33
img_temp = img_current # 1.if you use this
img_temp = copy.deepcopy(img_current) # 3.so we use the deepcopy img_temp = np.uint8(img_temp) img_temp[img_temp != label_id] = 0 # 2.there will cover the img_temp, and cover the img_current, in next loop, the data will be changed
in "mask2txt_custom.py" line 31-33