INTRO

I originally bought this projector for another project that required it to run 24 hours a day.

Unfortunately, the inexpensive projector failed much sooner than expected, forcing me to stop that project. Buying an expensive replacement wasn't an option, and I didn't want to create more electronic waste.

Before throwing it away, I opened it up and discovered that many of the components were still working. The main problem was the original control board.

So instead of throwing the whole projector away, I decided to build my own control system and reuse the parts that still worked.

I reused the original optical system and combined it with a Raspberry Pi 3, ILI9341 TFT display, camera, and computer vision.

The result is AI HoloProject X1 — an interactive projection system that turns an ordinary floor into a digital interface.

I didn't replace the projector. I replaced what was broken.

Project Goal

The goal was not simply to repair an old projector.

I wanted to give the hardware a completely new purpose: turning a normal floor into an interactive digital interface using projection and computer vision.

This project started with a broken projector.

It had originally been used for another project that required continuous operation, but the projector eventually failed.

Although the projector could no longer function normally, I wanted to find out whether its internal components could still be useful.

I opened the projector and inspected the internal components.

The main control board was no longer working, but several important components were still usable.

The optical system, projection lens, mechanical parts, cooling system, and other components could potentially be reused.

I carefully removed the optical components from the original projector.

The goal was to preserve as much of the original optical hardware as possible instead of replacing everything with new components.

The useful parts included the projection lens and optical elements used to guide and focus the image.

Since the original control board was no longer working, I decided to build my own control system.

This allowed me to reuse the working optical and mechanical components without having to purchase another complete projector.

The new system would provide the control and processing needed for the new application.

I integrated an ILI9341 TFT display into the modified optical system.

The display provides the graphical content that is processed through the optical system and projected onto the floor.

This allowed me to create a completely new image source while continuing to reuse the original projector optics.

A Raspberry Pi 3 is used as the main computer for the project.

It handles the graphical interface, camera input, interaction processing, and communication between the different parts of the system.

A projector can display an image, but it cannot understand what a user is doing.

To make the system interactive, I added a camera.

The camera observes the projected area and provides the input needed for computer vision and interaction detection.

A GUI Designed for the Floor

The next step was creating a simple graphical interface designed specifically for projection onto the floor.

Large, clearly separated buttons make the interface easier to see and easier for the camera to detect.

Example interface:

1. Video
2. Calendar

This circuit connects the ILI9341 TFT display to the Raspberry Pi 3 Model B while preserving the original pin configuration used in the project.

The ILI9341 provides the visual output, while the Raspberry Pi 3 handles the display control and processing. The circuit uses the required SPI communication lines, together with the control signals for the display.

The Raspberry Pi Camera is connected separately through the Pi 3's CSI camera interface. It is used as the visual input for the computer-vision interaction system.

The wiring shown in this diagram follows the actual circuit used in AI HoloProject X1.

Key connections:

1. ILI9341 → Raspberry Pi 3
2. MOSI (D10)
3. SCK (D9)
4. CE0 (D8)
5. D24
6. D25
7. Power and Ground
8. Raspberry Pi Camera → CSI interface

After completing the control system and software, I started rebuilding the projector using the original mechanical structure and the recovered optical components.

Instead of designing a completely new projector enclosure, I reused the original structure wherever possible.

The Raspberry Pi 3 and the new control hardware were integrated into the rebuilt system, while the original optical components remained part of the projection path.

This allowed the original projector to become the mechanical foundation of the new AI HoloProject X1 system.

With the main structure rebuilt, I installed the ILI9341 display into the original optical system.

The display acts as the new image source for the projector.

The original projection optics were kept and reused, while the ILI9341 provides the image that is sent through the optical path.

This is the key transformation of the project: the original projector no longer depends on its failed control board. Instead, the recovered optical system is driven by the new electronics.

One of the most important parts of this project was reusing the original optical and mechanical components.

The projector was no longer useful as a complete device, but its optical system and mechanical parts were still valuable.

I kept these components and integrated them into the new system instead of replacing them with new projector hardware.

This approach reduced the amount of electronic waste and gave the original components a completely different purpose.

The Raspberry Pi 3 became the main computer of the rebuilt system.

It connects the graphical interface, camera input, display control, and interaction processing together.

The new electronics replace the failed original projector controller while allowing the recovered optical hardware to continue being used.

This created a completely new architecture around the original projector components.

After installing the display and rebuilding the optical system, I tested the optical path to make sure the image could pass through the original projector optics correctly.

The position of the display and optical components had to be adjusted so that the projected image was usable.

This step was important because I was not using a complete projector anymore. The original optical system had to work together with the new display and electronics.

After the basic software was working, I needed to calibrate the projected interactive area.

Because the projected image and camera view must match, I created a calibration interface to define the main Region of Interest (ROI) and the locations of the interactive buttons.

The calibration screen allows the position and size of each area to be adjusted before saving the final configuration.

This makes the computer vision system more reliable when detecting interaction within the projected interface.

After calibration, I tested the graphical interface with different pages and interactive areas.

The Raspberry Pi displays the interface while the camera system monitors the defined regions.

The interface was designed with large and clearly separated buttons so that the projected elements can be detected more easily by the computer vision system.

I then tested the computer vision system using the camera view.

The software identifies the main interaction area and the individual button regions inside it.

The colored bounding boxes provide a visual way to verify that the detection areas are correctly positioned before using the system for real interaction.

After completing the software and calibration tests, I tested the system using the actual projection setup.

The image generated by the ILI9341 is passed through the recovered optical system and projected onto the working area.

At this stage, the system is no longer just a software prototype. The new electronics, display, camera, and recovered projector optics are working together as a complete system.