如何从浮点数4得知像素颜色是否为黑色

Question

I ask my question how to know pixel color using t_sampler in jocl in a different way I want to konw if a pixel is black or white knowing that I am using t_sampler in my kernel

const sampler_t smp = CLK_NORMALIZED_COORDS_FALSE | //Natural coordinates
                      CLK_ADDRESS_CLAMP | //Clamp to zeros
                      CLK_FILTER_NEAREST; //Don't interpolate

Then I used

int2 coord = (int2)(get_global_id(0), get_global_id(1));
    float4 pixel = read_imageui(input, smp, coord);

My question is: how to use the value pixel to know the color of the concerned pixel?

I am stuck from few days and I tried many solutions to solve this problem, If you need any clarifications I will respond. Here is my kernel code

const sampler_t smp = CLK_NORMALIZED_COORDS_FALSE | //Natural coordinates
                      CLK_ADDRESS_CLAMP | //Clamp to zeros
                      CLK_FILTER_NEAREST; //Don't interpolate
__kernel void basic(__read_only image2d_t input,__global float *result) 
{

   int gidX = get_global_id(0);
    int gidY = get_global_id(1);

int2 coord = (int2)(get_global_id(0), get_global_id(1));

int2 posIn = {gidX, gidY};

      float4 pixel = read_imagef(input, smp, posIn);
if ((pixel.x==0.0) && (pixel.y==0.0) && (pixel.z==0.0) ){

result[gidX]=1;
}  else result[gidX]=0;
    }

Answer 1

Read the pixel using read_imagef (I recommend it instead of read_imageui , depending on the image format, especially since in your example you are assigning to a float4 variable; check the specification for which is appropriate for each image format). The float4 components returned can then be checked. pixel.x is the red, pixel.y is the green, pixel.z is the blue, and pixel.w is the alpha (1.0f if the source format doesn't have alpha). To check for black (as you ask in the subject of your question), just check that .x and .y and .z are all 0.0.