Computer Vision at Seeed - Latest News from Seeed Studio

Computer Vision

Computer vision involves equipping computers with the ability to understand and interpret visual data from images and videos. This empowers them to identify objects, understand scenes, and even recognize actions. The applications span a wide spectrum, including crucial domains like medical diagnostics, self-driving cars, surveillance, and immersive augmented reality experiences. These capabilities are the result of significant strides in artificial intelligence, particularly in the realm of deep learning techniques.

Robotic Automation

Robotic automation involves using robots to independently perform tasks, often replacing human labor. These robots are equipped with sensors and programming to adapt to their surroundings. This technology improves efficiency, safety, and consistency in various industries like manufacturing and logistics, offering benefits like cost reduction and increased productivity. AI and machine learning enable robots to learn and enhance their performance over time.

Agriculture and Animal Husbandry

Agriculture and animal husbandry are more crucial than ever. The integration of computer vision brings transformative changes to these sectors. By merging computer vision with agriculture, we achieve smart crop monitoring, disease detection, and efficient resource usage. Similarly, in animal husbandry, computer vision enhances health monitoring, environment assessment, and behavior analysis. These advancements invigorate both industries, offering precise and sustainable management tools.

Object/Movement Detection

Object and movement detection involve using technology to identify and track objects and their motions in images or videos. This is done through algorithms and machine learning, which are used to spot specific objects or movements. This technology has uses in surveillance, self-driving cars, robotics, and video analysis, helping systems understand and react to the changing environment.

Home Automation Safety

Through computer vision in home automation, we achieve smart environmental monitoring, device control, and energy management. This tech enables devices to autonomously adjust for comfort and energy efficiency. Regarding security, real-time activity monitoring and anomaly detection offer crucial alerts for family safety and property protection. Computer vision revolutionizes home living with innovation in automation and security.

Users' Applications

Robot Iroi and Patrovor Integrated with 1:N Simultaneous Monitoring for Security

Dogugonggan provides customizable security AI and simultaneous monitoring solutions tailored to the environment and client’s needs. Dogugonggan uses robots to perform simple and repetitive tasks of patrol guards, with its autonomous driving algorithm that allows the robot to operate both indoors and outdoors, improving the working environment and ensuring security.

Edge AI-Enabled Ball Retriever Robot for Tennis Game

Tired of wasting time chasing stray tennis balls? Vive Robotics delivered an amazing ball-retrieving solution powered by the great AI inferencing capability of NVIDIA Jetson Nano, which could save you up to 20% of your on-court playing time and energy. By outsourcing ball retrieval to AI-powered robots, players can enjoy continuous rallies and a more streamlined game experience.

Rough Terrain Robot for Farm & Construction Site Deployment

As a pioneer in the field of rugged terrain, CuboRex launched the innovative tracked Robot for farm use, which is powered by NVIDIA Jetson Orin NX edge device to deal with the complex and difficult field working environment to reduce the labor of farmers based on the actual experience of working in the field. The robot kit also allows AI/CV engineers to start developing autonomous driving applications as soon as the robot arrives.

Deploy YOLOv8 on NVIDIA Jetson using TensorRT

This wiki guide explains how to deploy a YOLOv8 model into NVIDIA Jetson Platform and perform inference using TensorRT. Here we use TensorRT to maximize the inference performance on the Jetson platform. Different computer vision tasks will be introduced here such as:

- Object Detection/Tracking
- Image Segmentation/Classification
- Pose Estimation

Automated Pizza Making System with Consistent High-Quality Food Processing

GOPIZZA introduces an innovative pizza production system that decreases labor expenses and ensures consistent high-quality food. By utilizing the powerful NVIDIA Jetson Nano developer kit, the whole system simplifies operational management, reducing the number of required employees from four to at least one depending on store size. With state-of-the-art computer vision and edge AI inferencing capability, managing your store operations has never been easier.

MoveNet and reTerminal for Real-Time Human Activity Recognition

Traditionally, such tasks were performed on powerful servers or cloud-based platforms, often posing challenges related to privacy, latency, and resource constraints. In this blog, we delve into the importance of identifying human activities using video classification at the edge, exploring the benefits, applications, and the potential impact on various industries.

RocketCam with Seeed Xiao ESP32S3 Sense

The project “RocketCam” uses Seeed Xiao ESP32-S3 Sense board to create a small camera system. It’s designed to capture images from a model rocket’s viewpoint during its flight. The ESP32-S3 Sense board and camera module transmit real-time images to a user’s device through WiFi, accessible via a custom mobile app. The link provides detailed instructions for the project setup and construction.

TinyML Gesture Recognition demo

SenseCAP A1101 – LoRaWAN Vision AI Sensor is a TinyML Edge AI-enabled image sensor. It supports training models with TensorFlow Lite without writing a line of code, enabling developers or enthusiasts without advanced AI training experience like myself to learn AI machine learning and train AI models for interesting AI applications.

AI-driven LoRaWAN Fertilizer Pollution Detector

After producing organic fertilizers in different decomposition stages, I applied them to the soil in three separate flowerpots. Then, I added chemical fertilizers to each flowerpot in the same amount to examine the excessive use of chemical fertilizers depending on the soil integrity and structure. To demonstrate the fertilizer contamination effects on the environment, I sowed different types of tomato seedlings in each flowerpot.

Animal Recognition Practices on Grove Vision AI

At present, there is not much public research on the work edge AI for wildlife or animals. One of the few publicly available animal datasets – Animals Detection Images Dataset from Kaggle (called the “animals-80” dataset ) has been used. It contains 80 animals in 9.6GB images and should be great enough for common animal recognition tasks.

Creating a Smart and Safe Haven

In this article, we’ll delve into the dynamic synergy between the Seeed Studio reTerminal,

Home Assistant OS, and the Frigate add-on with Coral AI USB TPU. This transformative trio offers a unique opportunity to establish a centralized, privacy-conscious, and AI-driven home automation setup that not only enhances your lifestyle but also ensures the safety of your loved ones and property.

Utilizing Raspberry Pi-powered HMI reTerminal for Fall Detection System across Industries

This project is beneficial for elderly or disabled individuals who may be prone to falls, as well as their caregivers who can receive alerts and provide immediate assistance. It can also be utilized in industries such as healthcare, nursing homes, and assisted living facilities to improve safety and response times in case of falls.

Computer Vision

Users' Applications

Robot Iroi and Patrovor Integrated with 1:N Simultaneous Monitoring for Security

Edge AI-Enabled Ball Retriever Robot for Tennis Game

Rough Terrain Robot for Farm & Construction Site Deployment

Deploy YOLOv8 on NVIDIA Jetson using TensorRT

Automated Pizza Making System with Consistent High-Quality Food Processing

MoveNet and reTerminal for Real-Time Human Activity Recognition

RocketCam with Seeed Xiao ESP32S3 Sense

TinyML Gesture Recognition demo

AI-driven LoRaWAN Fertilizer Pollution Detector

Animal Recognition Practices on Grove Vision AI

Creating a Smart and Safe Haven

Utilizing Raspberry Pi-powered HMI reTerminal for Fall Detection System across Industries

Featured Devices

reTerminal DM (Camera & PoE)

XIAO ESP32S3 Sense

Grove - Vision AI Module

NVIDIA® Jetson Orin™ Nano Developer Kit

reTerminal CM4104032

SenseCAP K1101 - IoT Course for beginners

reComputer J4012