image processing

**matsp** · 10-14-2007

Yes, that looks about right.

--
Mats

**matsp** · 10-14-2007

Originally Posted by mike_g

You seem to have mixed up blue and green. Colour elements tend to go RGB.

This is far from guaranteed. Traditionally, Windows draws in "BGR" at the hardware level - it may well represent the image as RGB in user-mode, but once it gets to the driver it will be BGR. [Don't ask me where this comes from - but the hardware I worked on had a bit to say "RGB" or "BGR" when you configure the chip - and it looks "real funny" if you get it wrong].

There's obviously no importance to the order of the colours if we are inverting them - all colours are treated equally. If we were to do things like colour adjustment, we obviously need to understand which colour is which.

--
Mats

**taurus** · 10-14-2007

Originally Posted by matsp

Yes, that looks about right.

--
Mats

actually shouldnt it be data(ep100_lib_get_data)

?

**matsp** · 10-14-2007

Originally Posted by taurus

actually shouldnt it be data(ep100_lib_get_data)

?

No, ep100_lib_get_data() is a function, and data is, in what you've posted earlier, an array.

--
Mats

**taurus** · 10-14-2007

quick question, i cant really understand the difference between these 2, they sound the same:
// This routine will put the image pixel data into the supplied matrix
// from top left to bottom right. You must ensure that enough space is
// available.
void ep100_lib_get_data(int data[EP100_LIB_MAX_X][EP100_LIB_MAX_Y]);

// This routine will set the image pixel data from the supplied matrix
// from top left to bottom right. You must ensure that enough space is
// available.
void ep100_lib_set_data(int data[EP100_LIB_MAX_X][EP100_LIB_MAX_Y]);

**matsp** · 10-14-2007

Originally Posted by taurus

quick question, i cant really understand the difference between these 2, they sound the same:
// This routine will put the image pixel data into the supplied matrix
// from top left to bottom right. You must ensure that enough space is
// available.
void ep100_lib_get_data(int data[EP100_LIB_MAX_X][EP100_LIB_MAX_Y]);

// This routine will set the image pixel data from the supplied matrix
// from top left to bottom right. You must ensure that enough space is
// available.
void ep100_lib_set_data(int data[EP100_LIB_MAX_X][EP100_LIB_MAX_Y]);

Red words should make it clear which does what - no, they don't do the same thing, just like getchar() and putchar() don't do the same thing.

--
Mats

**taurus** · 10-14-2007

ok iv done some work, hopefully iv understood it, so could someone have a look through this pseudocode and advice further(it has left out somethings):

Code:

// Invert the image (1 = first argument after a.out)
IF argv[1] = -i THEN
INTEGER data[EP100_LIB_MAX_X][EP100_LIB_MAX_Y]		
ep100_lib_get_data(data)
ep100_lib_print_data(data)

FUNCTION invert_image(data)

ep100_lib_set_data(data) // Display original image

SLEEP  5 // Pauses program for 5 seconds

	ep100_lib_display_image(); // Display inverted image

Code:

ALGORITHM invert_image

INTEGER FUNCTION invert_image(INTEGER data[EP100_LIB_MAX_X] [EP100_LIB_MAX_Y])
FOR X  1 to EP100_LIB_MAX_X DO
	FOR Y  1 to EP100_LIB_MAX_Y DO
		INTEGER data[EP100_LIB_MAX_X] [EP100_LIB_MAX_Y]
		red  ((data[X][Y]>>16)&0xff)
		blue  ((data[X][Y]>>8)&0xff)
		green  ((data[X][Y])&0xff)
		
		data.red  255 – data.red
		data.blue  255 - data.blue
		data.green  255 - data.green
		
		RETURN 
	ENDFOR
ENDFOR

END ALGORITHM

THANKS

**matsp** · 10-14-2007

Like I said in a different post, you probably want to use the image's actual dimensions, as if you have a 100 x 100 pixel image, and your Max X/Y values are 1000, you'll process 990000 pixels out of a million worht of "emptyness", which is a bit of a waste, don't you think.

--
Mats

**taurus** · 10-14-2007

that ep100_lib_max x etc is defined as 1024
in the assignment it didnt say we get the dimensions from user, we just have to use CLA where a.out -i image.jpeg
-1 = invert

but otherwise does it look alright?

**ICool** · 10-14-2007

> SLEEP  5 // Pauses program for 5 seconds
I don't think you should be using SLEEP, I think I read it's not portable/ resource heavy ?

**matsp** · 10-14-2007

Code:

// This function returns the width in pixels of the loaded image.
int ep100_lib_get_image_width();

// This function returns the height in pixels of the loaded image.
int ep100_lib_get_image_height();

was posted by you in another thread. You don't think those functions would do exactly what you need to avoid making 99% of the work unnecessarily?

--
Mats

**taurus** · 10-14-2007

ohhhh ya k, ill change that, thanks for reminding me!

now its just how to scroll i have to work on

**matsp** · 10-14-2007

Originally Posted by ICool

> SLEEP  5 // Pauses program for 5 seconds
I don't think you should be using SLEEP, I think I read it's not portable/ resource heavy ?

Not to mention that it's completely meaningless. Why would you want to make it run slower than it can? [it won't be INSTANT anyways - although it shouldn't take more than fractin of a second.

If you want to PAUSE soemthing, then you ask the user to press enter or some such - that way, the pause is the right length for the user to see whatever is happening. Waiting X seconds just gets annoying.

--
Mats

**ICool** · 10-14-2007

I as thinking about scrolling :
Here for this example

Code:

1 2 3 4 5 
1 2 3 4 5
1 2 3 4 5

To get it scrolled to this :

Code:

4 5 1 2 3
4 5 1 2 3
4 5 1 2 3

This being a horizontal scroll I assume I need to work out where there is a "split" line , dump the contents either side of the split line into 2 arrays , and then rejoin the arrays , swapping their order . Would this be a good way to do it ?