What's wrong with this RGB to XYZ color space conversion algorithm?

Question

My goal is to convert an RGB pixel into CIELab color space for some special calculations only possible in CIELab. For this, I must convert RGB to XYZ first, which is the rea

Answer 1

As they said:

var_R = ( R / 255 )  ->  var_R = ( R / 255.0 ) or var_R = ( R * 0.003922 ) 
var_G = ( G / 255 )  ->  var_G = ( G / 255.0 ) or var_G = ( G * 0.003922 )
var_B = ( B / 255 )  ->  var_B = ( B / 255.0 ) or var_B = ( B * 0.003922 )

This is because of implicit conversion. Despite the fact that var_R, var_G and var_B variables are float type, the / operator sees two integers R, G, B and 255. it has to divide and returns an integer.

In order to get a float type value you can do a CAST or convert at least one of the variables into float type adding of a decimal point as follows:

var_B = ( B / 255.0f)

Another example of RGB2LAB and LAB2RGB conversion (bear in mind that is CIE Lab D65):

void RGB2LAB(uint8_t R, uint8_t G, uint8_t B, float *l, float *a, float *b) {
    float RGB[3], XYZ[3];

    RGB[0] = R * 0.003922;
    RGB[1] = G * 0.003922;
    RGB[2] = B * 0.003922;

    RGB[0] = (RGB[0] > 0.04045) ? pow(((RGB[0] + 0.055)/1.055), 2.4) : RGB[0] / 12.92;
    RGB[1] = (RGB[1] > 0.04045) ? pow(((RGB[1] + 0.055)/1.055), 2.4) : RGB[1] / 12.92;
    RGB[2] = (RGB[2] > 0.04045) ? pow(((RGB[2] + 0.055)/1.055), 2.4) : RGB[2] / 12.92;

    XYZ[0] = 0.412424  * RGB[0] + 0.357579 * RGB[1] + 0.180464  * RGB[2];
    XYZ[1] = 0.212656  * RGB[0] + 0.715158 * RGB[1] + 0.0721856 * RGB[2];
    XYZ[2] = 0.0193324 * RGB[0] + 0.119193 * RGB[1] + 0.950444  * RGB[2];

    *l = 116 * ( ( XYZ[1] / 1.000000) > 0.008856 ? pow(XYZ[1] / 1.000000, 0.333333) : 7.787 * XYZ[1] / 1.000000 + 0.137931) - 16;
    *a = 500 * ( ((XYZ[0] / 0.950467) > 0.008856 ? pow(XYZ[0] / 0.950467, 0.333333) : 7.787 * XYZ[0] / 0.950467 + 0.137931) - ((XYZ[1] / 1.000000) > 0.008856 ? pow(XYZ[1] / 1.000000, 0.333333) : 7.787 * XYZ[1] / 1.000000 + 0.137931) );
    *b = 200 * ( ((XYZ[1] / 1.000000) > 0.008856 ? pow(XYZ[1] / 1.000000, 0.333333) : 7.787 * XYZ[1] / 1.000000 + 0.137931) - ((XYZ[2] / 1.088969) > 0.008856 ? pow(XYZ[2] / 1.088969, 0.333333) : 7.787 * XYZ[2] / 1.088969 + 0.137931) );
}

void LAB2RGB(float L, float A, float B, uint8_t *r, uint8_t *g, uint8_t *b) {
    float XYZ[3], RGB[3];

    XYZ[1] = (L + 16 ) / 116;
    XYZ[0] = A / 500 + XYZ[1];
    XYZ[2] = XYZ[1] - B / 200;

    XYZ[1] = (XYZ[1]*XYZ[1]*XYZ[1] > 0.008856) ? XYZ[1]*XYZ[1]*XYZ[1] : (XYZ[1] - (16 / 116)) / 7.787;
    XYZ[0] = (XYZ[0]*XYZ[0]*XYZ[0] > 0.008856) ? XYZ[0]*XYZ[0]*XYZ[0] : (XYZ[0] - (16 / 116)) / 7.787;
    XYZ[2] = (XYZ[2]*XYZ[2]*XYZ[2] > 0.008856) ? XYZ[2]*XYZ[2]*XYZ[2] : (XYZ[2] - (16 / 116)) / 7.787;

    RGB[0] = 0.950467 * XYZ[0] *  3.2406 + 1.000000 * XYZ[1] * -1.5372 + 1.088969 * XYZ[2] * -0.4986;
    RGB[1] = 0.950467 * XYZ[0] * -0.9689 + 1.000000 * XYZ[1] *  1.8758 + 1.088969 * XYZ[2] *  0.0415;
    RGB[2] = 0.950467 * XYZ[0] *  0.0557 + 1.000000 * XYZ[1] * -0.2040 + 1.088969 * XYZ[2] *  1.0570;

    *r = (255 * ( (RGB[0] > 0.0031308) ? 1.055 * (pow(RGB[0], (1/2.4)) - 0.055) : RGB[0] * 12.92 ));
    *g = (255 * ( (RGB[1] > 0.0031308) ? 1.055 * (pow(RGB[1], (1/2.4)) - 0.055) : RGB[1] * 12.92 ));
    *b = (255 * ( (RGB[2] > 0.0031308) ? 1.055 * (pow(RGB[2], (1/2.4)) - 0.055) : RGB[2] * 12.92 ));
}