Introduction: DIY a Pocket Monsters Discriminator

About: Empowering Creation for Future Innovators; Our mission is to form a community with easy access to whether hardware, software and ideas that allow makers and younger generation to achieve their goals and realiz…

Recently, I heard that HuskyLens have released a new function: object classification.

Hardly have time to check it, today I decided to investigate it. Read through the full post, I have to say the object classification of HuskyLens is really powerful. But what to do now? There are too many choices, how about posting it to the other day? Before being beat by procrastination and becoming a couch potato, I find my notebook on the desk.

Hit it off! As the new function is object classification, how about make a discriminator to identify pocket monsters?

Due to time limitation, I do not make the external structure

of the discriminator, but only took advantage of the micro: bit wireless transmission function and the mind+ real-time mode, and make a simple simulation. Here's a brief explanation of the design process.


Micro: bit x2
HuskyLens x1(update to V0.4.9class)
micro: bit Driver Expansion Board x1

Step 1: Mainboard

Although there is no special requirement of mainboards for HuskyLens, the real-time mode of mind+ does not support add the HuskyLens extension. So, we have to take advantage of wireless communication, connecting the transmitter to a HuskyLens, then receiver to the real-time mode of mind+. Hereafter, micro: bit realizes the wireless transmission easily, without the need to transition through the Internet of Things platform.

Step 2: HuskyLens Learning

Just search pocket monsters online, then pointing HuskyLens to these images to learn.
When HuskyLens is learning, you can long-press the learning button and move the camera back and forth for multi-angle learning, it is recommended to learn more than 30 pictures at a time Oh, after learning 1 4 seconds press the learning key to continue learning. It is recommended to learn 30 images at least. Learning each picture one by one, and the switch time should be within 4s.

Step 3: Code for the Transmitter

The program logic is simple, when HuskyLens identifies different ID numbers, it sends the corresponding number to the micro: bit via the wireless communication. Remember to eliminate jitters.

Step 4: Code for the Receiver

It is simple to realize. We just need to add micro: bit and broadcast function in extensions. When the receiver receives data from the transmitter, according to the different numbers, its broadcast function will trigger to display different images.

Step 5: Spirits of Pokemons

I got the spirits of pokemon from

Thanks for reading.