(Not an expert at any of this, just someone with a bit of amateur experience):
Regular graphic design is a more artistic than engineering/scientific discipline. But any digital camera is basically a, err, "image microphone". You just capture and edit frequencies of light instead of sound.
If you get one that shoots RAW (basically any prosumer mirrorless, SLR, or compact would do that, as will many cell phones), you can bypass much of the on-board processing and tinker with it yourself in Lightroom or Photoshop or the manufacturer's own software. You set your own exposure levels (gain), saturation, noise reduction, white balance, etc. Usually these are presented to the user as simple sliders that obscure the underlying signals processing, but I guess you can say the same thing about something like Audacity. Except, of course, the preview is a picture instead of a sound clip.
Low/high pass filters in photography are typically manipulated using curves instead (https://photographyhero.com/understanding-tone-curve-lightro....), but that histogram is basically a graphical equalizer.
HDR is literally that: high dynamic range, where you exercise more control over the range compression (the camera captures more than regular JPEG can display) and use different algorithms to merge the highlights and shadows together "artistically" (or gaudily, often).
Some monitors and image codecs can also display real HDR, but IME that's more common in the film and video world. And a lot of Photoshop filters are themselves just signals algorithms, including the various blurs (and because of that, they can sometimes be reversed).
But just shoot some pictures and play with it in a photo editor. Creative Cloud monthly is a cheap way to play around with it without spending hundreds of dollars.
Fun intersection: You can also generate spectrograms that are also images, because it's all just signal: https://ificl.github.io/images-that-sound/ (and of course there's standard steganography, where you can embed arbitrary data inside an image)
-------------
If you're more interested in the signals part of it, there are fields who use similar tools to do what you're talking about: image-based signals processing. GIS (geographic information systems) especially comes to mind, where you're using signals processing and statistical techniques (and these days, machine learning) to analyze aerial and satellite imagery taken in different bands of data. The commercial sector uses ESRI for this; the open-source stuff is under https://grass.osgeo.org/ and https://www.qgis.org/. A lot of GIS analyses is on directly on raster image data, ingesting sensor data (from CCDs and the like) into bitmaps that they then run various filters on top of, and then comb through either with human eyes or algorithms of various sorts. Just some random examples... using deep learning to identify solar panels (https://storymaps.arcgis.com/stories/b9de83bf40644200a946ec5...), or this one with more detail (https://arxiv.org/pdf/1607.06029).
I think but am not sure that astronomy also use some similar techniques, sometimes with visible light, sometimes with RF or infrared, often to generate composites for both scientific use and public consumption.
Combining those techniques get you things like terahertz imagers at airport security, CT and MRI scanners, through-the-wall stud imagers, etc. Many of the techniques are similar from sound to visual to RF.
More generally, any sort of machine vision or non-visible-light imaging will also use those same techniques.
Or if you're more interested in the encoding of signal into files, image codecs (JPEG, PNG, WebP, JPEG2000, JPEGXL, AVIF, etc.) is a fascinating subfield full of people way smarter than me.
--------------
Hope some of that helps? If there's some particular part of this you're interested in, can you share more details?