Skip to content
Kaito Snezhko
Sign in
← all projects
● in progress2026

Image processing with hand-made data structures

University assignment in C++ — load BMP images and process every pixel through self-written stacks, queues and linked row lists, without the STL and without memory leaks.

role
Student (Data Structures, Universidad de Alcalá)
stack
C++17 · CodeLite · Make

Task

Lab assignment 1 of Data Structures (Universidad de Alcalá, fall 2026). Read a BMP image, store its pixels in linked data structures I wrote myself, transform the image in several stages and write the results back as BMP files.

What's done

  • Six small classes: Pixel, StackDS, QueueDS, RowListDS, ImageListDS and Core.
  • Stage 1: ask for an image (and ask again if it can't be loaded), then push every pixel onto the stack.
  • Stage 2: write the image back from the stack. A helper stack restores the original, so size and top are unchanged.
  • I found and fixed a row-padding bug in the provided BMP library: images whose width isn't a multiple of 4 came out broken.

Quality

  • The round trip is identical to the original, pixel by pixel (tested from 12 to 640 000 pixels).
  • 0 memory leaks and 0 warnings with -Wall -Wextra.

The repository is private while the course is running. The ideas are written up in the know-how article.