P-07activeEmbedded Vision · 2026
Bekkusu-CAM — Retro ESP32-CAM Web Camera
A standalone ESP32-CAM digital camera with a retro local web UI, client-side image processing, hardware shutter input, and dual stylized output modes.
- Role
- Developer · Embedded Vision & Web Interface
- Timeframe
- Mar 2026
- Domain
- Embedded Vision
- Status
- active
- Board
- AI Thinker ESP32-CAM
- Sensor
- OV2640
- Modes
- JPN / VTG
Overview
Bekkusu-CAM turns an AI-Thinker ESP32-CAM module into a compact standalone camera with its own WiFi hotspot and a browser-based control panel.
The system serves its UI directly from the microcontroller and applies retro image processing in the browser for a low-latency, self-hosted capture workflow.
The Challenge
The main challenge was making a tiny camera feel complete: stable capture, responsive UI, fresh frames, and a usable photo workflow without relying on cloud services or external apps.
Approach
- 01Built the camera as a self-hosted access point using ESP32-CAM AP mode.
- 02Served the UI locally and handled processing through client-side JavaScript canvas filters.
- 03Added physical shutter support through GPIO 13 with status polling for UI sync.
- 04Implemented dual aesthetic modes, JPN duotone and VTG vintage film, with timestamped downloads.
Outcomes
- Delivered a fully standalone camera experience with local-only operation.
- Enabled both browser control and hardware shutter capture.
- Created a distinctive retro-style output pipeline with two visual modes.
Stack
Hardware
- AI Thinker ESP32-CAM
- OV2640
- GPIO 4 Flash LED
- GPIO 13 Button
Firmware
- C++
- Arduino IDE / PlatformIO
- SPIFFS
Interface
- WiFi AP Mode
- HTML/CSS/JS
- Canvas Processing