Tested on version current version at the time (downloaded , not the online/browser version) , 2.4.1 

Selected Stable Diffusion for the model (in the 3rd button)

Enabled the img2img function (the 2nd button i mean), uploaded the following picture i wanted as a basis and set the denoising to 0.75 , no control net is selected : 

For prompt (1st button) i used 

high quality portrait by Jeremy Lipking

for negative i used

worst quality, ugly, low res, bad proportions

Then i generated 4 images.
Result all the image obtained are the same : 

Regenerating more times only result in always the same images being obtained.

I tested on another Stable Horde UI ( https://aqualxx.github.io/stable-ui/ ) and using the same setting, image basis and prompts it works without problem, i obtain images created by the AI using the original image as a basis, but on Lucid Creation it does nothing and only result in the original image unmodified .

I’ll take a look

I think i found the source of the broken img2img problem :

I always work with PNG because it has near lossless compression unlike JPG so when you paint or draw something in an art program the PNG does not degrade your work.
But this time i decided to convert my hand drawn image to JPG and try on Lucid Creation the same image i posted up there (but converted to JPG) and this time it worked : 

I then tried again with the PNG version and ended with the same problem as in the 1st post so it looks like unlike the other stable horde UI, Lucid Creation img2img does not work with PNG image, it will only work with JPG.

Make sure it doesn’t have an alpha channel