Newly upgraded and User-friendly AI Kit 2023

Introduction

Today we are introducing AI Kit(Artificial Intelligence Kit) 2023, an upgraded version of AI Kit. AI Kit 2023 is an entry-level AI Kit that combines visual, positioning, grabbing, and automatic sorting modules. The kit is based on the Python platform. It helps users who are just starting to learn the basics of AI and inspires them to think innovatively and understand the creative culture of open source.

AI Kit 2023 is a robotic arm application, simulating an industrial scenario. In this simulated industrial scenario, the robotic arm will replace manual labor to achieve smart sorting and implement preliminary industrial automation. The core of this scenario is the 5 visual algorithms and robotic arm motion control algorithms, which can quickly recognize objects and perform precise sorting.

The 5 vision algorithms are：

● Shape recognition

● Feature point recognition

● ArUco code recognition

● Color recognition

● YOLOv5 recognition

OpenCV is an open-source computer vision library for image processing, computer vision and machine learning. OpenCV supports multiple programming languages including C++ and Python, with a large number of image processing and computer vision functions to help developers implement various applications, such as image recognition, object tracking, robot navigation, etc.

The first 4 algorithms are for image processing and machine vision based on OpenCV algorithms, including color space recognition, feature point recognition, ArUco code recognition, and shape recognition.

The YOLOv5 (You Only Look Once version 5) is a popular object detection algorithm in recent years, v5 is the latest version. It uses Convolutional Neural Networks (CNNs) to predict objects in images and can perform detection quickly while still maintaining high accuracy.

Let's take a look at how the robotic arm works on AI Kit 2023.

Shape recognition algorithm

Feature point recognition algorithm

Color recognition algorithm

QR code recognition algorithm

YOLOv5 recognition algorithm

Next, we will show you how to use it.

Process

Environment

The AI Kit 2023 program is a suite of programs written in Python.

For using the robotic arms in M5Stack(myCobot, mechArm, myPalletizer), need to install the Python compilation environment and some corresponding Python libraries computer or PC, such as openCV.

For Raspberry Pi version robotic arms, there's no need to install any environment and it can be used right after power is on.

The program used by AI Kit 2023 needs to be downloaded from Github.

https://github.com/elephantrobotics/AiKit_UI

Let's get the program started!

run the main.py

This is the AI Kit 2023 visualisation program interface

Here we introduce the functions of this interface.

This function establishes communication between the program and the arm, selecting the arm's model, serial number and baud rate for connection.

Here is the live feed from the camera is displayed to process the relevant algorithms for machine vision.

This is the most crucial part of the program, where different algorithms are selected in the algorithm selection field to perform different functions. This is the general explanation of what this visual interface is all about.

Make sure you select the right algorithm.

Step:

Connect the robotic arm->Select the algorithm->Open the camera->Enable program Auto mode

This is the end of the basic introduction to the operation.

Summary

How is the performance of this upgraded AI kit?

As mentioned earlier, this is a kit that is very suitable for beginners to learn machine vision algorithms, robotic arm basics, and the use of python.

The entire program is open source, and each piece of code is written with detailed comments, allowing users to freely modify the code. From this, users can understand how a program framework is built and most importantly, how to integrate visual algorithms with robotic arm motion control. There are not too many complicated parameters and everything is simple and straightforward.

When it comes to robotic arm applications, "hand-eye calibration" cannot be separated. Hand-eye calibration is very important in robotic arm applications and there are two situations: eye to hand* and eye in hand*, to deal with the relationship between real-world data and computer data.

For those who have AI Kit, they will instantly know what makes AI Kit 2023 special.

The UI is much user-friendly since it's visualized, and it is designed to display data, status, results, etc., providing an intuitive and easy-to-understand information display. The visualization interface simplifies operations and helps users quickly identify, access, and operate data, and can interact and operate in multiple ways. Even if you don't understand the code at all, you can still successfully get the entire kit running normally.

Many users probably want to know about the upgrades of AI Kit 2023 compared to previous AI Kits, what are the differences. We will explain this to you in future articles! If you like this article, feel free to leave a comment and like it, see you next time.

*eye to hand:The camera is fixed in one place and is not affected by the movement of the robotic arm.

*eye in hand:The camera is on the robotic arm and follows the movement of the robotic arm