HLS based Deep Neural Network Accelerator Library for Xilinx Ultrascale+ MPSoCs

Smart camera with OV 7670 and Zynq

A minimalist Deep Learning framework for embedded Computer Vision

Graduation project repository, Real-time vehicle detection using two different approaches. HOG+SVM traditional approach and Deep Learning based approach using state of the art YOLO convolutional neural network.

Embedded Vision for Baby Behavior Monitoring in IoT

MIPI CSI-2 kernel driver for Onsemi AR0234 global shutter sensor on NVIDIA Jetson Orin NX / Nano (JetPack 6)

Embedded Vision for MVS in IoT

Traffic Violation Detection System is a system that utilize machine learning and embedded device to detect a traffic violation especially on light. The system would/should notify the authorities if it detect an object inside caution area.

Utilizing SEEED Studio Grove Vision AI V2 Module

Modular C++ OpenCV/GStreamer pipeline for Jetson, Raspberry Pi, and Linux edge devices.

GStreamer plugins for zero-copy NVMM video on NVIDIA Jetson (Xavier, Orin). Allocator, VIC transform, shared-memory IPC.

NVIDIA DeepStream GStreamer plugin for GPU-accelerated AprilTag detection on Jetson Orin NX/AGX using CUDA. Based on apriltags_cuda by FRC Team 766. Supports 8 simultaneous RTSP streams with <20% CPU usage.

A compact autonomous robot using Raspberry Pi integrates SLAM, YOLOv5 for real-time human detection, and RRT* for optimal path planning. It maps surroundings, detects a person as a goal, and navigates safely. Built in Python, it's lightweight and ideal for rescue, assistance, and indoor navigation.

An open and practical guide to Edge AI Engineering.

Official Project Page: "Efficient Self-Supervised Neuro-Analytic Visual Servoing for Real-time Quadrotor Control"

Repository is part of my BSc thesis with title "Gesture recognition in video streams on an embedded device"

Embedded Vision project which enables real-time control of RGB LED colors and brightness using hand gesture recognition. It combines Computer Vision (OpenCV + MediaPipe) and STM32 microcontroller using PWM on LEDS.

An open and practical guide to Edge Vision.

Real-time 720×480 @ 60 FPS grayscale vision pipeline — Teensy 4.x + MT9V034 + FlexIO2 + DMA. Feasibility study for ASIC design.

A high-performance Red Color Segmentation system using classic Computer Vision (OpenCV) on Raspberry Pi. Features dual-range HSV thresholding and morphological refinement, tested across multi-domain datasets including automotive and botanical subjects