Help on converting dubble to uint8 ?

2016 9 18
1 个回答
回答已采纳
更新时间：2016 9 18
4 次查看（30 天）

0 个投票

Hey,
I have wave signal that change between 2-20 and it's type of dubble. I need to convert it to type uint8, for example:
The wave signale is INDX(change from 2 to 20) type of dubble. Convert signal is: ADR = uint8(INDX);
the result of that conversion gives me values from 0 to 20. Even when ADR is runs from 0 to 18 or 1 to 19 I get the same result : ADR = 0 to 20 ?
Is it possible to get the same values of INDX after using INDX input ? for example, if INDX is running from 3 to 17, the result (ADR=uint8(INDX)) will be 3 to 17 instead 0f 0 to 20 ? or, if INDX is running from 2 to 18, the result (ADR=uint8(INDX)) will be 2 to 18 instead 0f 0 to 20 ? or, if INDX is running from 2 to 20, the result (ADR=uint8(INDX)) will be 2 to 20 instead 0f 0 to 20 ?
I need to keep the same values of the start and end(of INDX) after using uint8(INDX). uint8 type ,must be use.
Thanks, Henry

请先登录，再回答此问题。

 采纳的回答

Walter Roberson
Walter Roberson 2016-9-18

0 个投票

You must have made a mistake somewhere.
>> uint8(3:17)
ans =
3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
In the expression 3:17 the values are double precision values, so this shows that converting double precision values to uint8 works as expected for small non-negative values.

5 个评论
显示 3更早的评论 隐藏 3更早的评论

Henry Buck
Henry Buck 2016-9-18
编辑：Stephen23 2016-9-18
Hi,
well, that's the problem: the values of INDX are not a vector as you wrote.
INDX = round(A/B). B is a constant - let say 40. A is kind a sine wave - max value is 800 and min value is 0 - that gives INDX is between 0 to 20 In other case it couild be 90 to 700 - that gives INDX between 2 to 18..
whent I am using uint8(INDX), it gave me 21 values - 0 to 20.
Why is that ?
thanks, Henry
Stephen23
Stephen23 2016-9-18
@Henry Buck: please show us exactly the 21 values with this behavior, and how you are converting them to uint8. Then we can explain what is happening.
Image Analyst
Image Analyst 2016-9-18
编辑：Image Analyst 2016-9-18
Is it possible you use im2double() somewhere along the way? That might rescale and shift things. gray2mat() also scales and shifts to the 0 to 1 range. Also, uint8 rounds internally. It does not just chop off the fractional part.
Like everyone else is saying, you need to show more code, or at least a small similar example, like Walter did, that proves what you're saying.
Walter Roberson
Walter Roberson 2016-9-18
Not being a vector is irrelevant.
>> A = 53*(reshape(3:17, 3, 5) + rand(3,5)); B = 48; INDX = uint8(A/B)
INDX =
4 8 10 14 17
5 9 12 15 19
6 9 13 16 19
Henry Buck
Henry Buck 2016-9-18
hey guys,
I have to be honest. It is my mistake.
Thanks again.
Henry

请先登录，再进行评论。

更多回答（0 个）

类别

在 帮助中心 和 File Exchange 中查找有关 Data Type Conversion 的更多信息

标签

尚未输入任何标签。

Community Treasure Hunt

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

Start Hunting!

Translated by