[展翅高飛]

圖像可以分爲 bitmap 和vector 兩种， bitmap 就是用像素，而vector應該就是用綫條的。我個人認爲有了vector相機后，dc 便沒有了向像素的分別，因爲vector基本上可以無限放大。
但是我有三個問題：
1. vector 相機技術上可不可行？
2. vector 相片的size會有多大呢？
3. 這技術會不會引發其它問題？

小弟只是一個無知的攝影初哥，請各位多多指教。^^

[Tsun~~]

但是用線條會唔會無哂細節?

[展翅高飛]

Thank you for all the useful information, i really did learn a lot of stuff!

小弟只是一個無知的攝影初哥，請各位多多指教。^^

[vaworld]

Do u know using vector to represent raster images is much larger?

Old postscript printers were very slow in printing raster image for they need to convert raster to vecters. Today postscript printers have native processing in raster images like jpeg, tiff, etc.

[suloea]

me thinks it'll be rather difficult for the processor to take a photo in vector, considering the nature of the current ccd/cmos and even the basis of camera.
if one of such camera does exist, the pixel race may eventually end...

[hippo]

CCD / CMOS 記錄資料, 是用一點點的. 要轉成 vector, 是否不太合適?

[展翅高飛]

I am a student, but i am not researching this. I just have this question for ages, then i hope some of you could answer me, Thank you very much for all the answers and knowledge ^^

小弟只是一個無知的攝影初哥，請各位多多指教。^^

[展翅高飛]

Out of interest, D2c, are you a professor of somewhere? You seemed so professional and know so much about these imaging techniques.

[suloea]

//quote
Your idea is to replicate unlimited info, out of limit info, right???

In info theory, that's impossible, unless the info is of infinitely repeating patterns.

Digital Zoom

- extra-polate infinite info out of limited info

- but we see the poor output quality already (everybody wants an optical instead of digital zoom DC).
//endquote

thing about vector is it doesn't replicate info when resizing.
it is the info providing the constuction of a picture(direction and magnitude), so when u acutally resizing a vector image u multiply the magnitude, not replicates or extrapolating.
an easy example can be obtained on google where u try to look for a vector picutre, and resizing it with photoshop or illustrator.

thats just my 2 cents though

[展翅高飛]

thank you^^

[hlleung8]

可能叮噹個八寶袋裏面有 =P

呢條 thread 好 academic, 小弟都開了眼界 =)

[suloea]

wow
thanks for clearing up that for me
im just some year 1 cs student and didn't really know those stuffs in depth
think ill get there one day ;P
and honestly, that ain't my field of studies anyway
its just so happened that i started toying with cameras this summer and did some reading on the net



btw, if u had your msc oversea, why bother working in hk anyway?

[by1989]

我想這是不可能的吧 @@"
vector圖像不錯是可以任意resize...
首先, 點陣圖 -> 向量圖暫時係未見得可以實現...
既使實現左...同直接拉到張jpeg無分別...細節一樣唔會出現...就係因為咁...所以要高質量可以resize既圖像...都唔會將佢地轉成vector(或者跟本轉唔到)
ex:
|.| << 真實環境(有一點)
| | << 點陣圖
轉換成向量後 | | >> 拉大
| |
| | < 果一點e然唔會出現

再去諗...要用幾多條公式先可以真真正正代表真實世界既一樣野 ?? 接近無限...
如果要用有限既公式去代表應用無限既公式去代表既野...細節就會lose...要解決...就要鬥用得多公式...就好似上面個例子咁...真實世界到...2條線入面一定有野...再加多條...咁果3條線中間e然會有野...你要表達哂兩條線既野...就要用接近無限條線...但係咁樣係無可能...結果只可以用有限既線代表無限既野...既然愈多線就可以表達到中間既細節愈多...咁樣...各家都會盡力去整可以處理多d公式既ccd/cmos同處理器...咁同pixel既competition有咩分別?

[by1989]

你要表達哂兩條線既野...就要用接近無限條線
更正為
你要表達哂兩條線之間既野...就要用接近無限條公式

[vaworld]

老實講句 vector camera 會唔會有d係本末倒至?

因為影完相之後先用 software 將個raster image convert 去 vector重實際.

因為相機本身都要咁樣做先可以產生 vector.

無端端要部相機做多一大串之後都做得的東西, 無形中減低部相機性能.