清空筛选条件
清空筛选条件

How to Add Image Classes?

2 次查看（过去 30 天）
Qvna Lhyvnav
Qvna Lhyvnav 2019-5-18
评论： Qvna Lhyvnav 2019-5-18
I want to classify my image files into 4 classes, but I only get 0 and 1 with this code:
Target_Data = ones(1,Total_Images);
Target_Data(1:Total_Images/4) = 0;
How to add the other classes? Thank you.

请先登录，再回答此问题。

采纳的回答

Walter Roberson
Walter Roberson 2019-5-18
Target_Data = repmat((1:4).', Total_Images/4, 1);
  7 个评论
显示 5更早的评论
Walter Roberson
Walter Roberson 2019-5-18
Target_Data = repelem(1:4, Total_Images/4);
Qvna Lhyvnav
Qvna Lhyvnav 2019-5-18
Done. Thank you for your help!

请先登录，再进行评论。

更多回答（1 个）

KALYAN ACHARJYA
KALYAN ACHARJYA 2019-5-18
编辑：KALYAN ACHARJYA 2019-5-18
May be you are confusing with logical indexing, there is nothing related to classification of images
lets say Total_Images=8, that menas Target_Data= ones(1,8)
>> Target_Data=ones(1,Total_Images)
Target_Data =
1 1 1 1 1 1 1 1
in next line Total_Images/4=2, so the next line is Target_Data(1:2)=0
>> Target_Data(1:Total_Images/4)=0
Target_Data =
0 0 1 1 1 1 1 1
It is checking the 1 D array elements from 1 to 2 (Target_Data(1:2)) equal to zero or not, its false logic, therefore it is zero,
There for 1st and 2nd elements are 0 and rest are same as initial Taret__Data.
If you do following one, lets check all Target_Data elemets, wheather 0 or not, you may get all zero in resultant target elements.
>> Target_Data(1:8)=0
Target_Data =
0 0 0 0 0 0 0 0
I didnot find any clue to classes images based on these two statements, if you can, as you now undestand the code, do modification accordingly.
  1 个评论
Qvna Lhyvnav
Qvna Lhyvnav 2019-5-18
Thank you for your reply. I understand, but I want to make four classes (60 files with each 15 files has different values). This code works for two classes.

请先登录，再进行评论。

类别

在 Help Center 和 File Exchange 中查找有关 Image Processing and Computer Vision 的更多信息

标签

产品


版本

R2017a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by