41. Person Detection

41.1. Overview

The person detection example demos the use of a camera to capture an image, detect person in images and locate them.

  • Optimisation with the SIMD

  • Worke with the mcunet

41.2. Configurations

  • Connect the camera device to the CAM interface on the board

  • Click LCD-800480W070TC section and connect

41.3. Toolchain Requirements

This example requires a DSP-enabled toolchain to utilize hardware DSP instructions for neural network acceleration:

  • IDE: Segger Embedded Studio (or other IDEs with DSP toolchain support)

  • GCC Toolchain: Andes RISC-V toolchain with DSP extension support or ZCC toolchain with DSP extension support

Note

Standard GCC toolchains without DSP extensions will not be able to compile and run this demo correctly. The DSP instructions are essential for efficient neural network inference.

41.4. Code Generation

Following the steps in this readme will allow you to generate code in the codegen module. After generating the code, you can find the codegen directory in the source code directory.

41.5. Code Options

  • To get faster program running, you need modify the link file, place the data and bss segments into the RAM near the CPU. If there is not enough space, ensure that the arrays in genModel.h are located in the RAM near the CPU.

41.6. Running the example

When the example runs successfully, The LCD will display the live screen, the processed image and the recognition result, serial output:

person object example
Get: class 0: 1
(332, 796, 127, 465)
fps:5.846605.
Get: class 0: 2
(419, 799, 29, 476)
(5, 785, 0, 479)
fps:5.866579.
Get: class 0: 2
(442, 777, 44, 479)
(5, 785, 0, 479)
fps:5.873189