[Erha Image Recognition Artificial Intelligence Vision Sensor] Evaluation 1: Winner Chip Level Evaluation

mameng

[Erha Image Recognition Artificial Intelligence Vision Sensor] Evaluation 1: Winner Chip Level Evaluation [Copy link]

 

I am very happy to receive the Erha Image Recognition AI Vision Sensor Development Kit from E-World and DFRobo Community. DFRobo is headquartered in Pudong, Shanghai, and is an internationally leading high-tech enterprise engaged in open source hardware, robotics products, STEM education, maker education and artificial intelligence education. Since its establishment, DFRobot has launched thousands of self-developed products and provided maker education products and solutions to educational and training institutions around the world. It has currently provided maker education products to more than 20,000 teachers and students in more than 8,000 schools in more than 200 countries and regions around the world.

Empower Creation is our core mission. Helping the creators of current and future human society to create better is our greatest value.

— Ricky Ye CEO of DFRobot

Gravity: HuskyLens AI Vision Sensor:

HuskyLens uses a new generation of AI chip Kendryte K210, with a built-in 64-bit 400MHz dual-core RISC-V processor. Kendryte K210 is a system-on-chip (SoC) that integrates machine vision and machine hearing. It uses TSMC's ultra-low power 28nm advanced process and a dual-core 64-bit processor, which has better power consumption performance, stability and reliability. The program strives for zero threshold development and can be deployed to users' products in the shortest time, bringing artificial intelligence to the product. Kendryte K210 is located in the SoC of the artificial intelligence and Internet of Things market and is a very convenient single-chip microcomputer. Kendryte means investigation intelligence in Chinese, and investigation intelligence comes from investigation and search. The main application field of this chip is the Internet of Things field, and it was developed in the field of Internet of Things. Therefore, this chip mainly provides artificial intelligence solutions and is a research hotspot in the field of artificial intelligence. The classic YOLO artificial intelligence algorithm can only run at a speed of 1~2 frames per second on STM32H743, but it can be more than 30 times faster on HuskyLens. It performs exceptionally well, even when capturing fast-moving objects.

HuskyLens is an easy-to-use artificial intelligence visual sensor with 6 built-in functions: face recognition, object tracking, object recognition, line tracking, color recognition, and label recognition. AI training can be completed with just one button, getting rid of tedious training and complex visual algorithms, and focusing more on the conception and implementation of the project. HuskyLens has an onboard UART / I2C interface that can be connected to mainstream controllers such as Arduino, Raspberry Pi, LattePanda, micro:bit, etc. It comes with a 2.0-inch IPS display. When adjusting parameters, there is no need for computer assistance. The debugging process and recognition results are directly displayed on the screen. What you see is what you get, which is very convenient. It realizes seamless hardware docking and directly outputs recognition results to the controller. You can create very creative projects without having to deal with complex algorithms.

The function button is a three-way review button, which can be pressed left, right, up and down, and a learning button.

Built-in firmware boot animation: Erha appears!

Run image face recognition, YYDS!

soso

Looking forward to the follow-up :)