How to embed image in the whitespace between words

Question

I would like to have an image 'embedded' in a page within my document, but done so with the whitespace between words.

Ultimately I would like to have a moving image if you flick through the pages.

I am surprised by the absolute nothing that I have come up with!! A solution involving a monochrome image would be ideal.

Just a little something for my avid fans 8)

Answer 1

Having taken a bit of Q&A to understand your intent, I post this "answer" both to clarify for other readers what I believe the intent of your question to be, plus to assert that I think it cannot be done with any success. In my MWE, I try to place a simple circle in the rivers of text. To help show that circle to the readers, I present again the result with the rivers of whitespace replaced with red dashes. So why don't I think it feasible?

1) In my MWE, the rivers are exaggerated with multiple spaces, yet still are virtually impossible to discern, even when highlighted with red dashes. To accomplish it without such flagrant exaggeration is infinitely more difficult;

2) While whitespace rivers can be a distraction, I've never heard it asserted that they could be the focus of one's observation;

3) the "pixel resolution" of whitespace is so coarse, that even over the course of full page, the resolution is not great enough to represent an image in the white space as anything other than an unresolved blob, in my opinion.

But by all means, someone please prove me wrong.

\documentclass{article}
\usepackage{verbatimbox}
\usepackage{xcolor}
\begin{document}
\let\svdash-
\catcode`-=\active
\def\coloron{\def-{\textcolor{red}{\svdash}}}
\begin{verbnobox}[\rmfamily\coloron]
Here is our goal.  It is a test.  What
we are trying     to see is whether or
if an    image can arise   in
the    rivers of  this text.   That
is   to say,  can one see the    circle
that   is formed  in large    rivers of 
my pic?    Maybe if you    squint, one
can just make        it out. Barely.
Then again, maybe not.
~
Below, the relevant rivers are replaced with dashes
~
Here is our goal.  It is a test.  What
we are trying-----to see is whether or
if an----image can arise---in
the----rivers of  this text.---That
is---to say,  can one see the----circle
that---is formed  in large----rivers of 
my pic?----Maybe if you----squint, one
can just make--------it out. Barely.
Then again, maybe not.
\end{verbnobox}
\end{document}

---

If you really wanted to embed a hidden image, and I suggest a more mature message than that contemplated by your labmate "friends", there are much easier ways:

\documentclass{article}
\usepackage{stackengine,xcolor, graphicx}
\begin{document}
Where is the hidden image%
\stackinset{c}{-.2pt}{b}{.3pt}{\scalebox{0.02}{\textcolor{white}{Hi, mom}}}{?}
\end{document}

Answer 2

It can mostly be done 8) manually adjusting right margin is a PITA and not very accurate, so for educational purposes here is the finished ouptut which both proves and disproves that this can be done and that it works 8)

  \documentclass{article}
  \usepackage{verbatimbox}
  \usepackage{xcolor}
  \begin{document}
  \begin{verbnobox}[\rmfamily]
  target  of  a given  family’s toxin.   Many type II  TAs  target  translational  machinery,
  which  is  highly  conserved  across  all domains of life  (Poole  and  Logan 2005;  Noller
  2004).  DNA synthesis machinery, DNA polymerases  and  primases  are  less  conserved
  (Poole  and  Logan  2005;  Leipe, Aravind,  and  Koonin 1999;  Werner  and  Grohmann
  2011;  Aravind and  Koonin 2001).  Amongst type I  TAs, SymE,  which targets  mRNA
  (Kawano,  Aravind,  and Storz  2007),  would  provide  an  interesting comparison to the
  membrane proteins investigated within this thesis.

Clea   rly, neither toxi   n norantitoxin target  accounts  for  all  differences within  TA
  syste   ms, because families w   ith the same target often contain loci that differ inability
  to c   onfer a PSK effect on a    plasmid  (Szekeres et  al . 2007;  Christensen,  Maenhaut   
  Mich   el, et al. 2004; De Bast,    Mine, and  Van  Melderen  2008;  Wilbaux et al.  2007;
  Fiebig    et al. 2010). Similarly, ty   pe I  and  III  RMs  are not  as mobile and have not
  been sho   wn to induce PSK (Nadere   r et al.  2002;  O’Sullivan  et al.  2000;  Mruk and
  Kobayashi 2014),    but have the same tar   get (DNA) as  type  II  RMs.  Other  factors
  are also important for    PSK, including the le   vels at  which  the  toxin  and  antitoxin
  are expressed, and the rate    at which the toxin and    antitoxin are degraded in the cell
  after plasmid loss (Chapter 5).

TA systems rely on differential d   ecay of  the  toxin  and ant   itoxin  to  induce  PSK
  (Chapter 5). Type II  systems  are ti   ghtly  regulated  by  protein     antitoxins,   which
  interact with the toxin and act as transcr   iptional  repressors,  sensitive    to changes  in
  stoichiometry (Mruk and Kobayashi 2014;    J. Zhang,  Y.  Zhang, and M. In   ouye 2003;
  Kedzierska,  Lian,  and F.  Hayes  2007;    Cataudella, Sneppen, et al. 2013; Cata   udella,
  Trusina,  et  al.  2012; Afif  et al. 2001).    I used  equations  of  logarithmic decay    as a
  starting point for analyzing the condit   ions necessary for type  II  TAs to exhibit P   SK.
  In particular, I analyzed the pop   ulation of toxin and antitoxin in the cell necessary    for
  PSK given their respective half   -lives ( Chapter  5 ). Antitoxins  with similar stabil   ities
  as their associated toxins cannot    be  expressed  in  numbers  significantly  higher    than
  the toxin,  or  there  will  not be su   fficient free  toxin  to create a  PSK effect    within a
  given time period.  Given the  inherent    noise of ge   ne  expression (Rase   r and  O’Shea
  2005; C. V. Rao, D. M. Wolf, and Arkin 2002),    system   s relying    on stable antitoxins
  risk  plasmid  suicide from  excess toxin  in  the cell     prior to  loss. Thus, there  is likely
  to be an evolutionary trend toward TA systems with high expression levels of very
  \end{verbnobox}
  \end{document}