From: priyanka Tiwari on 24 Apr 2010 08:00 sir, i am working on 'detection of tuberoclosis bacili using digital image processing'. i am using colour segmentation for the extraction of bacili but i am facing problem in setting a threshold value .since the intensity of the colour in the whole bacili is not constant & it is changing abruptly in some areas. i've applied contrast enhancement algorithm as well for enhacing the characteristics of the bacili but it is not showing very good results. sir plz help me to solve this problem. priyanka
From: ImageAnalyst on 24 Apr 2010 08:36 priyanka: Without you posting a set of images showing the differing colors, there's nothing we can do. Post your images to http://drop.io if you want suggestions. Otherwise all I can say is to use color classification methods - like you're doing only more effectively.
From: priyanka Tiwari on 24 Apr 2010 09:00 sir' i've just sended you the images. sir i am also struggling with colour classification . but i need some guidance for the implementation of the colour classifcation. can i try another colour model like hsi or ycbcr ?? if you'll get the images plz reply. regards priyanka
From: priyanka Tiwari on 24 Apr 2010 09:09 link to images is - http://drop.io/tuberoclosis
|
Pages: 1 Prev: How to show the means and SEMs of two data series in one graph? Next: Help!!! |