How to extract an object from an image
19 visualizaciones (últimos 30 días)
Mostrar comentarios más antiguos
Nikhil Hegde
el 13 de Dic. de 2015
Respondida: Arnab Banerjee
el 13 de Sept. de 2019
I've got this code to extract a single cashew from a image containing multiple cashews. I need the extracted image to be RGB as well.
cashew_RGB = imread('White Wholes 180.jpg');
cashew_BW = img_BW(cashew_RGB);
cashew_stats = regionprops(bwlabel(cashew_BW),'BoundingBox');
cashew_single = imcrop(cashew_RGB,cashew_stats(i).BoundingBox);
I have attached the image and the user-defined function 'img_BW' which converts RGB image to black and white If you execute the code you'll find that the extracted cashew image has an invasion from the neighboring cashew. Is there any way to extract the cashew without such invasions? Thanks for your help!
0 comentarios
Respuesta aceptada
Image Analyst
el 13 de Dic. de 2015
You can't use the bounding box or else you may have incursions from another blob that juts into the bounding box. You have to use either PixelList from regionprops, or (easier) use ismember() with the blob you want to extract:
thisBlob = ismember(labeledImage, blobIndexToExtract) > 0;
6 comentarios
Jan
el 10 de En. de 2017
@nour: Please use flags for pointing out inproper use of the forum. Instead of attaching a new question as a comment, open a new thread. Thanks.
Más respuestas (2)
r r
el 8 de Ag. de 2018
I have an image I want to extract the small areas of black color within the blue How so؟؟؟
1 comentario
Arnab Banerjee
el 13 de Sept. de 2019
first, convert to gray, by rgb2gray(), then find the perfect threshold of the region from the image matrix. it can be aslo done in RGB image if you knew the R:G:B ration of the threshold.
lastly, past the bounded area in a mask, or extract the partcular threshold
0 comentarios
Ver también
Categorías
Más información sobre Convert Image Type en Help Center y File Exchange.
Productos
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!