how to choose which is best

2 次查看（过去 30 天）
FIR
FIR 2011-12-17
hi I have extracted the features of image such as colour ,shape and texture,no i have to apply a technique to select which is best from colour shape or texture,can anyone tell which technique or algorithm to apply please

请先登录，再回答此问题。

采纳的回答

Walter Roberson
Walter Roberson 2011-12-17
The most important technique is to mathematically define what you mean by "best" in this context. Without a mathematical definition for your situation, you are not going to be able to resolve this.
  5 个评论
显示 3更早的评论
Walter Roberson
Walter Roberson 2011-12-20
Did you create the requirement of "best", or did someone else? If you are the one who created that requirement, then change the requirements until you get something that you can mathematically define. If someone else created that requirement, go back to them and talk to them about what is required.
I have no algorithms to suggest to you: you have asked for something that does not have a meaning. Your question is like asking "What is best, pizza, beer, or chocolate?"
FIR
FIR 2011-12-20
ok walter i will come back to you after defining the meaning for best

请先登录，再进行评论。

更多回答（1 个）

Jan
Jan 2011-12-19
It is impossible to give you a valid suggestion. It depends on the input and the wanted output, but you did not define them exactly.
Currently your question is equivalent to: "I have different sets of numbers. Which one is the best?"
Using the Buddha Toolbox, the answer will be:
The one with the smallest information (maximum entropy), because there is no information at all.
  2 个评论
FIR
FIR 2011-12-20
Jan i have posted it ,ok ,in other way i have image features formats stores in cell in .mat file
[1x30 double] [1x60 double] [1x30 double]
[1x30 double] [1x30 double] [1x30 double]
[1x30 double] [1x30 double] [1x30 double]
from these is it able to find which is best,these are the output of three different images,where each row represents extracted features of colour,shape,texture
Jan
Jan 2011-12-20
@FIR: There is still no definition of "best". You have to start from a mathematical definition of "good". If you build the gradient over this expression, you will find the defition of "better". And at the points, where "better" is zero, you get either the "best" or "worst" images.

请先登录，再进行评论。

类别

在 Help Center 和 File Exchange 中查找有关 Image Category Classification 的更多信息

标签

Community Treasure Hunt

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

Start Hunting!

Translated by