Create program to assist the sorting of LEGO (Image Recognition, Data Entry)
Budget: $30 – $250 AUD
Not entirely sure what requires to be done to reach the goal, but I know what the goal is....
Each LEGO brick has its own unique identifier number, however, it takes a really long time to sort a container full of random LEGO bricks and catalog the parts.
I want to speed up the process of identifying LEGO bricks, by placing them under the camera, and the camera being able to discern which LEGO piece it is.
For a rough demonstration, look at: https://youtu.be/04JkdHEX3Yk?t=69
What the program needs to do:
Using a camera, the camera is to 'scan' a single lego piece travelling along a small white conveyer belt, which just spits out the brick into a container (the camera will have proper lighting and plain white background). With the images gathered by the camera, and using a "Convolutional Neural Network" or whatever works, the program is to compare the images to models found in the LDraw depository (https://www.ldraw.org/parts/direct-parts-access.html) , in order to find/estimate the brick's part number.
After the part number has been recognized, the program must type out the part number into an Excel spreadsheet, or if possible, make an entry into BrickStore (https://www.brickforge.de/brickstore/).
Would be happy to explain a bit better the ideal user process - just want to see how much this project would cost approximately.
Thanks!
Each LEGO brick has its own unique identifier number, however, it takes a really long time to sort a container full of random LEGO bricks and catalog the parts.
I want to speed up the process of identifying LEGO bricks, by placing them under the camera, and the camera being able to discern which LEGO piece it is.
For a rough demonstration, look at: https://youtu.be/04JkdHEX3Yk?t=69
What the program needs to do:
Using a camera, the camera is to 'scan' a single lego piece travelling along a small white conveyer belt, which just spits out the brick into a container (the camera will have proper lighting and plain white background). With the images gathered by the camera, and using a "Convolutional Neural Network" or whatever works, the program is to compare the images to models found in the LDraw depository (https://www.ldraw.org/parts/direct-parts-access.html) , in order to find/estimate the brick's part number.
After the part number has been recognized, the program must type out the part number into an Excel spreadsheet, or if possible, make an entry into BrickStore (https://www.brickforge.de/brickstore/).
Would be happy to explain a bit better the ideal user process - just want to see how much this project would cost approximately.
Thanks!