Transparency Maker reads in a .jpg or .png and creates a Pixel Information Database, that can be manipulated in a language I invented called BQL or Bitmap Query language.

The reason I wrote this program originally is when I purchase stock photos, almost all come with a background color and often I need a transparent background. There are other tools that perform background removal, with more now than in 2012 when I started working on Transparency Maker, however I wanted a way to programmaticaly manipulate pixels.

The original purpose of Transparency Maker was to make backgrounds transparent, thus the name. I have now expanded the features to include updating pixels, changing colors, etc.

What do you think? Should I keep the name Transparency Maker, or do you like Pixel Database better?

Let me know by creatiing an Issue on this project, or on any Transparency Maker videos on YouTube.

Visual Studio 2019 is the recommeneded version, but most of the code worked as far back as VS 2010.

I will make an install and put it on my website as soon as I get a chance. If anyone would like an install, create an issue on this project and / or leave a comment on any of my YouTube videos and I will expedite the install.

https://www.youtube.com/playlist?list=PLKrW5tXCPiX2PxrLPszDzlcEZwQG-Qb8r

I just updated my channel to be split into more play lists to separate each section.

The first thing that happens when an image is opened, is your file is parsed into a Pixel Information Database, aka Pixel Database.

When the program starts, click the Start button to select your image. The image must be a Png or a Jpg.

A Pixel Database is a List of PixelInformation objects that contain properties about the pixel. To find the pixel information for any pixel, turn the Color Picker on.

Tip: To determiine if the Color Picker is on, hover your mouse of any part of your image. If the mouse cursor turns into a pointer (a hand), then the Color Picker is on.

Once the Color Picker is turned on, click any pixel in your image and a Pixel Information Box pops up:

The box contains the following properties:

The Red value of RGB value. Value range is 0 - 255.

The Green value of RGB value. Value range is 0 - 255.

The Blue value of RGB value. Value range is 0 - 255.

The Transparency Level of the pixel. Values are 0 - 255, where 0 is transparent and 255 is visible.

The horizontal position of the pixel clicked.

The sum of Green plus Red.

The sum of Blue plus Green.

The sum of Blue plus Red.

The sum of Red plus Green plus Blue.

The vertical position of the pixel clicked.

You can use all of the above values in your queries to manipulate your image.

This application started out as a plain Windows Form that I would hard code C# statements to replace the background of stock photos with transparent backgrounds. Over time I switched to a more object oriented approach where I analyze the entire image, and thus the Pixel Database was created.

The initial criticism I have received is there are not any controls like most graphic programs. If that is what everyone still wants I am open to ideas about what types of controls are needed. I created BQL because it is similar to SQL that many C# developers already know, thus enabling them to get up to speed quickly. Another reason I created BQL instead of a control based graphical UI is time to develop. Many of my failed endeavors have involved very complicated UI's, and then I am reminded of this small company called Google that started with a webpage that contained a single textbox and a button.

There are two types of queries; Hide/Show and Update.

Any pixels affected by the query will be set to Alpha 0, which makes them invisible.

Hide Pixels Where
Total > 700

Hide Pixels Where can be shortened to:

Hide
Total > 700

Any pixels affected by the query will be set to Alpha 255, which makes them visible.

Show Pixels Where
Red < 200

Show Pixels Where can be shortened to:

Show
Red < 200

This query parser may be rewritten to be more robust someday, but for now certain attributes must be on their own line. I wrote the query parser in a few hours many years ago, and it works pretty well so I have not bothered to refactor it (yet).

So the query above would fail if written like this

Show X < 150

Show
X < 150

It wouldn't be that hard to rewrite the query parser, so let me know if you think the line by line requirement is confusing. I know it is crude, but this app was designed to be quick, dirty and functional at first.

Will match criteria on exact values.

Hide
Red = 233

All pixels that have a Red value of 233 will be set to Alpha 0 to be hidden.

Will match criterian on greater than or equal to

Show
Y > 1200

All pixels with a Y value of 1,200 or higher wil be set to Alpha 255 to be shown.

Will match criteria on less than or equal to

Hide
BlueGreen < 300

All pixels with a BlueGreen value of 300 or less wil be set to Alpha 0 to be hidden.

Will match criteria that is greater than or equal to the first number and less than or equal to the second number.

Hide
Blue Between 200 255

All pixels with a Blue value between 200 and 255 will be hidden.

You can combine criteria to further narrow down the pixels that are manipulated

Show
X > 200
Y Between 112 140
Total = 765

All pixels with X coordinate of 200 or higher and a Y value of 112 - 140 and a Total of 765 (White).

Update queries are very similar to Hide queries, except that you must include a Color attribute

There are two ways to set a color, Named Colors or RGB values.

You can set a pixel to a named color

Note: For a list of Dot Net Colors see System.Drawing.Colors or this website lists them: http://www.flounder.com/csharp_color_table.htm

Update
Set Color FireRed
Where
Total Between 125 150

You set the color by specifying the Red, Green and Blue values

Update
Set Color 121 220 7
Where
Y > 100

You must specify the Where, even if you want to update all pixels

Update
Set Color White
Where
Total > 0

One of the uses of Transparency Maker is an image that is grainy or pixelated, you can smooth out an area by setting all pixels in a range to a certain value.

Update
Set Color 220 0 55
Where
Total Between 525 590
X Between 200 360
Y > 400

To Draw a Transparent Line, Type The Following

Line 1: Draw Line LineThickness
Line 2: First two points are Line Start Point X Y
Second two points are Line End Point X Y
Line 3 (Optional): Repeat Direction Iterations Move

Draw Line 4
161 125 457 66
Repeat Down 15 1

Type Line 1 Draw Line (Thickness), then hit enter to place your cursor on the next line. Ensuriing the Color Picker is toggled on, Click on your image to establish Point 1:

Draw Line 4
161 125

Then click your mouse again for the end point of the line:

Draw Line 4
161 125 457 66

(Optional) Type line 3 if you want to repeat.

Then click Apply.

Draw Line 4
161 125 457 66
Repeat Down 15 1

Use with Repeat with Down, Up, Left and Right

Draw Line 2
0 64 457 66
Repeat Up 100 2

The above line will be drawn 100 times, and each time it will move the Y coordinate up (minus) two pixels.

Note: The line will go futher than the bounds of the image, so Repeat 50 would accomplish the same thing.

Draw Line 1
225 412 105 430
Repeat Left 25 2

The above line will be drawn 25 times and will move the StartPoint.X and EndPoint.X left two pixels each iteration.

Draw Line 5
140 300 90 600
Repeat Right 50 10

The above line will draw transparent stripes through an image because the Repeat Move (last parameter) is greater than the line thickness from Line 1.

Drawing transparent lines is slower than any other operation using Transparency Maker because what it actually does is create a copy of the source image and draws a line trhough the copy in a color that doesn't exist in the copy, then creates a Pixel Database out of the copy image and determines which pixels were modified. The pixels that were modified in the copy are then applied to the source Pixel Database and set to Transparent.

I think this is a good candidate for letting people with a powerful GPU take advantage of this, but I haven't learned how to apply this yet. I have a 1080 TI, but I have never written any code to take advantage of a GPU. Volunteers anyone?

Clicking Save sometimes causes an error. You can think click Continue, and are returned to your image. Click Save As works.

It has been on my list for a long time to add a Save prompt if there are changes and you attempt to close the file or the application.

Let me know what features you think are needed or any feedback you may have.

I thought I had code to draw a line in a specific color, but this has not been done yet apparently.

If anyone wants this feature let me know.

Any suggestions or feedback is welcome.