Mico python information for K230

Viewed 141
software micropython k230

First, I would like to express my gratitude for the support and guidance provided by this community. Your assistance has been invaluable in my journey so far.

I have a few questions regarding the SDKs, memory allocation, and customization of the MicroPython image for my CanMV-K230 image. Current img is CanMV-K230_V1P0_P1_micropython_v1.2.2-0-g4b8cae1_nncase_v2.9.0.img.

  1. The official repository describes three main SDKs: Linux-only, RTT-only, and RTT with Linux. I would like to know which of these SDKs is used to build the MicroPython images available for different boards.

  2. Since I am currently using the prebuilt image, flashing it onto an SD card, and programming on Windows, I want to understand the internal structure of this image. Specifically, I would like to know how memory is allocated across different sections and which core is responsible for each functionality.

  3. For customization, I am interested in knowing what parameters can be modified when building a custom image. I would also like to understand how memory allocation can be adjusted and how both cores can be utilized efficiently, as I believe my code is currently running only on the big core.

  4. Additionally, we have received new hardware, and I need to configure it correctly from the firmware side. I would appreciate guidance on key configurations and settings that should be adjusted when adapting the firmware to new hardware. (specially for canmv v1.1 and v3)

  5. One of the SDK examples demonstrates running an AI model on one core and the LVGL library on another. I am curious whether this approach is feasible in MicroPython and, if so, how it can be implemented.

  6. Is this possible to remove sd crad from final product after all r&d and store all data like kmodel and img to eMMC, i need to store face embedding bin file as well so how can i do that as well?

  7. Is this any necessity changes for camera(with and without isp) and display configuration to run on custom board?

I appreciate any insights the community can provide on these topics. Thank you in advance for your help!

Urvi_Zala1
1 Answers
  1. RTT-Only SDK is now build for micropython.
  2. the image generate configure file is here.
  3. you can follow the documents, download the sdk, run make menuconfig and make rtsmart-menuconfig, and can change the configure for the image.
  4. you can follow this documents to add a new board.
  5. now the micropython is not support very good for lvgl + ai, we will push a update about it.
  6. yes you can use eMMC instead of SDCard, our micropython sdk support it. just need change so configure.
  7. add support for camera and display need run make menuconfig to change the hardware configure of camera and display.
xelll1159