More Fake Pictures of Oswald

[Bill Chapman]

Well I'll just Swanee there Di Vinci. Aren't you clever? Obtain a hint on retouching.
So what we had... was a totally incompetent photographer and that shadow just happened to create a perfect line right down the neck. Also, that photograph was retouched and retouched again.
Then, it was colorized [retouching it once more]
Moving on with the Oswald photos....


 
 




Picasso out there? What do you say?...same guys? If so, Oswald looked younger in 1961 than he did in 1958.

Show us that you know anything about light & shade. And that you even know what I'm talking about.

[John Mytton]

Mytton's shtick is to morph two photos with each other and then just declare "clearly they are the same person".  Then morph two other photos with each other and then just declare "clearly they are not the same person".  Then claim that his "clear" opinions of the results are "scientific".  Keep in mind that every intermediate step in the morph is a completely fabricated image.

I don't declare anything, the images speak for themselves.   



In the following two morphs between two very close celebrity lookalikes and the original celebrities, the immediate obvious differences demonstrate just why this form of graphical comparison is so damn effective!

Keep in mind that every intermediate step in the morph is a completely fabricated image.

Exactly, that's the fundamental foundation for this advanced computerized imaging technique, the relative distance of each facial feature from the camera between two people in two photos are by definition mathematically identical therefore when the same person is morphed every consecutive frame reflects the fact that there is a direct relationship between each set of millions of individual pixels

The Apple iphone uses the exact same methodology, it takes a depth mapped photo of someone's face and stores this digital data for comparison with a live face, the same information can be unlocked by viewing two images stereoscopically or combined into easier viewing in a morph.



Much like a human fingerprint, our unique faces can be broken down into a complex 3D shape and from whatever angle all these individual features in 3D space will always be in the exact same ratio.





JohnM

[Jerry Freeman]

OK ...apparently, Oswald must have discovered Rogaine before Minoxidil was developed years later
Take a look again at the various hairlines and the cosmetic looking eyebrows.
Mo-jo motion ain't working.

[Jerry Freeman]

Show us that you know anything about light & shade. And that you even know what I'm talking about.

How can someone know what Chapman is 'talking about' when he doesn't even know himself?

[John Mytton]

OK ...apparently, Oswald must have discovered Rogaine before Minoxidil was developed years later
Take a look again at the various hairlines and the cosmetic looking eyebrows.
Mo-jo motion ain't working.

What the heck are you looking at, Oswald's dramatically receding hairline at the temple and resulting combover is clear in the arrest photo.





JohnM

[Jack Trojan]

Matt, this procedure is based on the same principle as the iphone facial recognition software and is virtually infallible.

JohnM

  This is absolute BS and confirms that you know squat about image analysis or facial recognition. Else pls explain how a "fade" effect is based on the same principle as facial recognition. This should be good.

Hint: Here's a fade algorithm with 14 lines of code.

    function getfadedcolor(Red1, Red2, Green1, Green2, Blue1, Blue2, fs, fn) {
        fa1 = (Red2 - Red1) / fs;
        fa2 = (Green2 - Green1) / fs;
        fa3 = (Blue2 - Blue1) / fs;
        cx1 = Red1;
        cx2 = Green1;
        cx3 = Blue1;
        for (var n = 1; n <= fn; n++) {
           cx1 += fa1;
           cx2 += fa2;
           cx3 += fa3;
        }
        return RGB(cx1, cx2, cx3);
    }

I used this simple algorithm to make the following animated gif (what gives with his body BTW?)



All fading/morphing algorithms are based on this simple methodology of interleaving images.

How many lines of code do you think are required for infallible facial recognition technology?

[Jerry Organ]

|   133-A   |   133-B   |   133-C   |

Oswald's left shoulder appears to be further back in space relative to the chest-midline in the 133-B picture than in the 122-A picture. This accounts for the difference in the shoulder width.

I also believe the morph...



... has the 133-B head a bit smaller than the 133-A head, which would in turn make the left shoulder more narrow to the camera. Not saying the animation sizing was intentional, but there's an amount of subjectivity that could induce error. Since the shadow-cast on the facial features in the two photos vary (due to head tilt), the width of the head might be a more accurate thing to align.

[John Mytton]

Hint: Here's a fade algorithm with 14 lines of code.

    function getfadedcolor(Red1, Red2, Green1, Green2, Blue1, Blue2, fs, fn) {
        fa1 = (Red2 - Red1) / fs;
        fa2 = (Green2 - Green1) / fs;
        fa3 = (Blue2 - Blue1) / fs;
        cx1 = Red1;
        cx2 = Green1;
        cx3 = Blue1;
        for (var n = 1; n <= fn; n++) {
           cx1 += fa1;
           cx2 += fa2;
           cx3 += fa3;
        }
        return RGB(cx1, cx2, cx3);
    }

I used this simple algorithm to make the following animated gif (what gives with his body BTW?)

I used this simple algorithm to make the following animated gif

 

The code you presented is a simple fade to white, how do you think that is relevant?
And please explain how your code applies to your gif which is just swapping one set of RGB values for another?



JohnM