目录
1、导出ONNX模型
from ultralytics import YOLO
# Load a model
model = YOLO(r'G:\Yolov8\yolov8-detect-pt\yolov8n.pt') # load an official model
# model = YOLO('path/to/best.pt') # load a custom trained
# Export the model
# model.export(format='onnx')
model.export(format='onnx')
运行得到onnx
验证导出的onnx是否可用。运行正常的话,保存的图片有检测出的物体。
from ultralytics import YOLO
import glob
import os
# Load a model
model = YOLO(r'G:\Yolov8\yolov8-detect-pt\yolov8n.onnx') # load an official model
# Predict with the model
imgpath = r'G:\Yolov8\ultralytics-main-detect\imgs'
imgs = glob.glob(os.path.join(imgpath,'*.jpg'))
for img in imgs:
model.predict(img, save=True)
2、在开发板中转为bin模型
(1)配置地平线提供的docker开发环境
本文在windows系统下配置的Docker环境。大家可自行配置。
(2)启动docker容器
如果你的电脑中以前没用过docker,需要先安装docker。
(3)onnx模型检查
进入到horizon_xj3_open_explorer_v1.8.5_20211224\ddk\samples\ai_toolchain\horizon_model_convert_sample\04_detection路径下,创建一个mapper文件夹(个人文件管理,可忽略),再创建一个08_yolov8文件夹(可自行命名), 在路径下创建onnx_model文件夹,将onnx放入
把在第一步得到的yolov8n.onnx模型,放到指定位置:
然后在mapper文件夹下,新建一个01_check.sh文件,内容如下:
#!/usr/bin/env sh
set -e -v
cd $(dirname $0) || exit
# 模型类型,本文以onnx为例
model_type="onnx"
# 要检查的onnx模型位置
onnx_model="./onnx_model/yolov8n.onnx"
# 检查输出日志,放到哪里去
# 虽然它还是放到了与01_check.sh同级目录下(感觉像小bug)
output="./model_output/yolov8_checker.log"
# 用的什么架构,不用改
march="bernoulli2"
hb_mapper checker --model-type ${model_type} \
--model ${onnx_model} \
--output ${output} --march ${march}
cd到mapper文件夹下,执行命令:
sh 01_check.sh
(4)处理图片
挑选一部分图片(20-100)放在mapper文件夹下
在mapper文件夹下,新建一个02_preprocess.sh文件,内容如下:
#!/usr/bin/env bash
# Copyright (c) 2020 Horizon Robotics.All Rights Reserved.
#
# The material in this file is confidential and contains trade secrets
# of Horizon Robotics Inc. This is proprietary information owned by
# Horizon Robotics Inc. No part of this work may be disclosed,
# reproduced, copied, transmitted, or used in any way for any purpose,
# without the express written permission of Horizon Robotics Inc.
set -e -v
cd $(dirname $0) || exit
python3 ../../../data_preprocess.py \
--src_dir data/pcd \
--dst_dir ./pcd_rgb_f32 \
--pic_ext .rgb \
--read_mode opencv
data_preprocess.py在下图位置(官方自带)
在mapper文件夹下,新建一个preprocess.py文件,内容如下:
target_size=(640, 640) -> 调整输出的size
# Copyright (c) 2021 Horizon Robotics.All Rights Reserved.
#
# The material in this file is confidential and contains trade secrets
# of Horizon Robotics Inc. This is proprietary information owned by
# Horizon Robotics Inc. No part of this work may be disclosed,
# reproduced, copied, transmitted, or used in any way for any purpose,
# without the express written permission of Horizon Robotics Inc.
import sys
sys.path.append("../../../01_common/python/data/")
from transformer import *
from dataloader import *
def calibration_transformers():
transformers = [
PadResizeTransformer(target_size=(640, 640)),
HWC2CHWTransformer(),
BGR2RGBTransformer(data_format="CHW"),
]
return transformers
def infer_transformers(input_shape, input_layout="NHWC"):
transformers = [
PadResizeTransformer(target_size=input_shape),
BGR2RGBTransformer(data_format="HWC"),
RGB2NV12Transformer(data_format="HWC"),
NV12ToYUV444Transformer(target_size=input_shape,
yuv444_output_layout=input_layout[1:]),
]
return transformers
def infer_image_preprocess(image_file, input_layout, input_shape):
transformers = infer_transformers(input_shape, input_layout)
origin_image, processed_image = SingleImageDataLoaderWithOrigin(
transformers, image_file, imread_mode="opencv")
return origin_image, processed_image
def eval_image_preprocess(image_path, annotation_path, input_shape,
input_layout):
transformers = infer_transformers(input_shape, input_layout)
data_loader = COCODataLoader(transformers,
image_path,
annotation_path,
imread_mode='opencv')
return data_loader
运行
sh 02_preprocess.sh
(5)转为.bin模型
准备好处理后的图像数据,下面就是获取能够在开发板上运行的模型了,地平线在开发板上运行的模型后缀为.bin,故在此称为.bin模型。
这一步需要准备两个文件,一个是03_build.sh,一个是yolov8_config.yaml,两个文件均放于mapper文件夹下。
注意:运行03_build.sh之前要保证前一步生成的rgb大小要和yolov8_config.yaml文件设置的输入尺寸一致(模型输入)。
03_build.sh
#!/bin/bash
set -e -v
cd $(dirname $0)
config_file="./yolov8_config.yaml"
model_type="onnx"
# build model
hb_mapper makertbin --config ${config_file} \
--model-type ${model_type}
yolov8_config.yaml
# Copyright (c) 2020 Horizon Robotics.All Rights Reserved.
#
# The material in this file is confidential and contains trade secrets
# of Horizon Robotics Inc. This is proprietary information owned by
# Horizon Robotics Inc. No part of this work may be disclosed,
# reproduced, copied, transmitted, or used in any way for any purpose,
# without the express written permission of Horizon Robotics Inc.
# 模型转化相关的参数
# ------------------------------------
# model conversion related parameters
model_parameters:
# Onnx浮点网络数据模型文件
# -----------------------------------------------------------
# the model file of floating-point ONNX neural network data
onnx_model: 'onnx_model/yolov8n.onnx'
# 适用BPU架构
# --------------------------------
# the applicable BPU architecture
march: "bernoulli2"
# 指定模型转换过程中是否输出各层的中间结果,如果为True,则输出所有层的中间输出结果,
# --------------------------------------------------------------------------------------
# specifies whether or not to dump the intermediate results of all layers in conversion
# if set to True, then the intermediate results of all layers shall be dumped
layer_out_dump: False
# 日志文件的输出控制参数,
# debug输出模型转换的详细信息
# info只输出关键信息
# warn输出警告和错误级别以上的信息
# ----------------------------------------------------------------------------------------
# output control parameter of log file(s),
# if set to 'debug', then details of model conversion will be dumped
# if set to 'info', then only important imformation will be dumped
# if set to 'warn', then information ranked higher than 'warn' and 'error' will be dumped
log_level: 'debug'
# 模型转换输出的结果的存放目录
# -----------------------------------------------------------
# the directory in which model conversion results are stored
working_dir: 'model_output'
# 模型转换输出的用于上板执行的模型文件的名称前缀
# -----------------------------------------------------------------------------------------
# model conversion generated name prefix of those model files used for dev board execution
output_model_file_prefix: 'yolov8n_rgb'
# 模型输入相关参数, 若输入多个节点, 则应使用';'进行分隔, 使用默认缺省设置则写None
# --------------------------------------------------------------------------
# model input related parameters,
# please use ";" to seperate when inputting multiple nodes,
# please use None for default setting
input_parameters:
# (选填) 模型输入的节点名称, 此名称应与模型文件中的名称一致, 否则会报错, 不填则会使用模型文件中的节点名称
# --------------------------------------------------------------------------------------------------------
# (Optional) node name of model input,
# it shall be the same as the name of model file, otherwise an error will be reported,
# the node name of model file will be used when left blank
input_name: ""
# 网络实际执行时,输入给网络的数据格式,包括 nv12/rgb/bgr/yuv444/gray/featuremap,
# ------------------------------------------------------------------------------------------
# the data formats to be passed into neural network when actually performing neural network
# available options: nv12/rgb/bgr/yuv444/gray/featuremap,
input_type_rt: 'nv12'
# 网络实际执行时输入的数据排布, 可选值为 NHWC/NCHW
# 若input_type_rt配置为nv12,则此处参数不需要配置
# ------------------------------------------------------------------
# the data layout formats to be passed into neural network when actually performing neural network, available options: NHWC/NCHW
# If input_type_rt is configured as nv12, then this parameter does not need to be configured
#input_layout_rt: ''
# 网络训练时输入的数据格式,可选的值为rgb/bgr/gray/featuremap/yuv444
# --------------------------------------------------------------------
# the data formats in network training
# available options: rgb/bgr/gray/featuremap/yuv444
input_type_train: 'rgb'
# 网络训练时输入的数据排布, 可选值为 NHWC/NCHW
# ------------------------------------------------------------------
# the data layout in network training, available options: NHWC/NCHW
input_layout_train: 'NCHW'
# (选填) 模型网络的输入大小, 以'x'分隔, 不填则会使用模型文件中的网络输入大小,否则会覆盖模型文件中输入大小
# -------------------------------------------------------------------------------------------
# (Optional)the input size of model network, seperated by 'x'
# note that the network input size of model file will be used if left blank
# otherwise it will overwrite the input size of model file
input_shape: '1x3x640x640'
# 网络实际执行时,输入给网络的batch_size, 默认值为1
# ---------------------------------------------------------------------
# the data batch_size to be passed into neural network when actually performing neural network, default value: 1
#input_batch: 1
# 网络输入的预处理方法,主要有以下几种:
# no_preprocess 不做任何操作
# data_mean 减去通道均值mean_value
# data_scale 对图像像素乘以data_scale系数
# data_mean_and_scale 减去通道均值后再乘以scale系数
# -------------------------------------------------------------------------------------------
# preprocessing methods of network input, available options:
# 'no_preprocess' indicates that no preprocess will be made
# 'data_mean' indicates that to minus the channel mean, i.e. mean_value
# 'data_scale' indicates that image pixels to multiply data_scale ratio
# 'data_mean_and_scale' indicates that to multiply scale ratio after channel mean is minused
norm_type: 'data_scale'
# 图像减去的均值, 如果是通道均值,value之间必须用空格分隔
# --------------------------------------------------------------------------
# the mean value minused by image
# note that values must be seperated by space if channel mean value is used
mean_value: ''
# 图像预处理缩放比例,如果是通道缩放比例,value之间必须用空格分隔
# ---------------------------------------------------------------------------
# scale value of image preprocess
# note that values must be seperated by space if channel scale value is used
scale_value: 0.003921568627451
# 模型量化相关参数
# -----------------------------
# model calibration parameters
calibration_parameters:
# 模型量化的参考图像的存放目录,图片格式支持Jpeg、Bmp等格式,输入的图片
# 应该是使用的典型场景,一般是从测试集中选择20~100张图片,另外输入
# 的图片要覆盖典型场景,不要是偏僻场景,如过曝光、饱和、模糊、纯黑、纯白等图片
# 若有多个输入节点, 则应使用';'进行分隔
# -------------------------------------------------------------------------------------------------
# the directory where reference images of model quantization are stored
# image formats include JPEG, BMP etc.
# should be classic application scenarios, usually 20~100 images are picked out from test datasets
# in addition, note that input images should cover typical scenarios
# and try to avoid those overexposed, oversaturated, vague,
# pure blank or pure white images
# use ';' to seperate when there are multiple input nodes
cal_data_dir: './pcd_rgb_f32'
# 如果输入的图片文件尺寸和模型训练的尺寸不一致时,并且preprocess_on为true,
# 则将采用默认预处理方法(skimage resize),
# 将输入图片缩放或者裁减到指定尺寸,否则,需要用户提前把图片处理为训练时的尺寸
# ---------------------------------------------------------------------------------
# In case the size of input image file is different from that of in model training
# and that preprocess_on is set to True,
# shall the default preprocess method(skimage resize) be used
# i.e., to resize or crop input image into specified size
# otherwise user must keep image size as that of in training in advance
preprocess_on: False
# 模型量化的算法类型,支持kl、max,通常采用KL即可满足要求, 若为QAT导出的模型, 则应选择load
# ----------------------------------------------------------------------------------
# types of model quantization algorithms, usually kl will meet the need
# available options:kl and max
# if converted model is quanti model exported from QAT , then choose `load`
calibration_type: 'default'
# 编译器相关参数
# ----------------------------
# compiler related parameters
compiler_parameters:
# 编译策略,支持bandwidth和latency两种优化模式;
# bandwidth以优化ddr的访问带宽为目标;
# latency以优化推理时间为目标
# -------------------------------------------------------------------------------------------
# compilation strategy, there are 2 available optimization modes: 'bandwidth' and 'lantency'
# the 'bandwidth' mode aims to optimize ddr access bandwidth
# while the 'lantency' mode aims to optimize inference duration
compile_mode: 'latency'
# 设置debug为True将打开编译器的debug模式,能够输出性能仿真的相关信息,如帧率、DDR带宽占用等
# -----------------------------------------------------------------------------------
# the compiler's debug mode will be enabled by setting to True
# this will dump performance simulation related information
# such as: frame rate, DDR bandwidth usage etc.
debug: False
# 编译模型指定核数,不指定默认编译单核模型, 若编译双核模型,将下边注释打开即可
# -------------------------------------------------------------------------------------
# specifies number of cores to be used in model compilation
# as default, single core is used as this value left blank
# please delete the "# " below to enable dual-core mode when compiling dual-core model
# core_num: 2
# 优化等级可选范围为O0~O3
# O0不做任何优化, 编译速度最快,优化程度最低,
# O1-O3随着优化等级提高,预期编译后的模型的执行速度会更快,但是所需编译时间也会变长。
# 推荐用O2做最快验证
# ----------------------------------------------------------------------------------------------------------
# optimization level ranges between O0~O3
# O0 indicates that no optimization will be made
# the faster the compilation, the lower optimization level will be
# O1-O3: as optimization levels increase gradually, model execution, after compilation, shall become faster
# while compilation will be prolonged
# it is recommended to use O2 for fastest verification
optimize_level: 'O3'
运行
sh 03_build.sh
转换成功后,得到几个文件:
3、验证模型
得到模型后,可自行编写后处理代码在地平线板子验证模型: