feat: expand image processing options

[LeviSnoot]

I thought the image processing could do with some user configuration and increased flexibility.

• Added the option to disable the blurred image background images.bg = false. This option keeps the 1:1 canvas, but has no background, makes it look more native to Discord:

• Add user config option to set a blur value for the blurred background images.bg_blur, default set to 3%.
• When images.bg is false, rounded corners are automatically applied to the image. The amount can be set using images.corner_radius. Default is 4%.

Some examples for various configurations:

Default:

{
  "images": {
    "bg": true,
    "bg_blur": 3
  }
}

Reduced image size:

{
  "images": {
    "size": 512,
    "bg": true,
    "bg_blur": 3
  }
}

Increased blur value:

{
  "images": {
    "size": 512,
    "bg": true,
    "bg_blur": 15
  }
}

Disabled background, default corner radius:

{
  "images": {
    "size": 512,
    "bg": false,
    "corner_radius": 4
  }
}

Increased corner radius:

{
  "images": {
    "size": 512,
    "bg": false,
    "corner_radius": 10
  }
}

No rounded corners:

{
  "images": {
    "size": 512,
    "bg": false,
    "corner_radius": 0
  }
}

[LeviSnoot]

@radiicall I believe the checks are only failing due to this: https://github.com/Radiicall/jellyfin-rpc/blob/92677b337cb80e2e882b7c9fb182e3d475b82720/jellyfin-rpc-cli/Cargo.toml#L32

I left it alone in this commit as I'm not sure if there's an external dependency on it, but when I uncomment this and compile it works just fine.