【剁手系列】Jetson Nano（四）跑通jetson-inference
这篇文章介绍怎么安装英伟达的AI学习库Jetson-Inference
（一）下载jetson-inference
剩下的过程就比较简单了。首先如果您没有安装git和cmake，先安装它们
sudo apt-get install git cmake
接着从git上克隆jetson-inference
git clone https://github.com/dusty-nv/jetson-inference
cd jetson-inference
git submodule update --init
（二）关键点
下面开始有分流了，有两种方法
第一种，直接可以下载网络模型的，可以直接配置cmake，选择你要下载的模型，以及是否安装pytorch等。
mkdir build
cd build
cmake ../
第二种，通过浏览器下载
进入tools这个目录，里面有download-models.sh这个文件，用gedit打开
根据函数里的网址，一个一个下好，我是在jetson-inference目录下新建了downloads这个文件夹，并把下载好的文件放在这个文件夹下面
然后修改刚刚打开的文件，找到attempt_download_file和attempt_download_archive这两个函数进行修改，修改完代码如下：
function attempt_download_file()
{
local filename=$1
if [ -e ../downloads/$filename ]
then
local wget_status=0
else
local wget_status=1
fi
local wget_status=$?
if [ $wget_status != 0 ]; then
echo "$LOG No such file '$filename' (error code=$wget_status)"
return $wget_status
fi
mv ../downloads/$filename $OUTPUT_DIR
return 0
}
function attempt_download_archive()
{
local filename=$1
if [ -e ../downloads/$filename ]
then
local wget_status=0
else
local wget_status=1
fi
if [ $wget_status != 0 ]; then
echo "$LOG No such file '$filename' (error code=$wget_status)"
return $wget_status
fi
tar -xzf ../downloads/$filename -C $OUTPUT_DIR
local tar_status=$?
if [ $tar_status != 0 ]; then
echo "$LOG tar failed to extract '$filename' (error code=$tar_status)"
return $tar_status
fi
rm ../downloads/$filename
return 0
}
对于pytorch，则是另外一个install-pytorch.sh文件，找到网址，将whl下载下来，自己安装一下就行了，到时候在选择安装pytorch的界面别选就行了。
最后还是老套路，cmake，记住上面提到的事项。
mkdir build
cd build
cmake ../
cmake成功后，就需要编译了
cd jetson-inference/build
make
sudo make install
（三）测试
cd jetson-inference/build/aarch64/bin
./detectnet-console peds_0.jpg output_0.jpg
结果如下
detectNet -- loading detection network model from:
-- model networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff
-- input_blob 'Input'
-- output_blob 'NMS'
-- output_count 'NMS_1'
-- class_labels networks/SSD-Mobilenet-v2/ssd_coco_labels.txt
-- threshold 0.500000
-- batch_size 1
[TRT] TensorRT version 5.1.6
[TRT] loading NVIDIA plugins...
[TRT] Plugin Creator registration succeeded - GridAnchor_TRT
[TRT] Plugin Creator registration succeeded - NMS_TRT
[TRT] Plugin Creator registration succeeded - Reorg_TRT
[TRT] Plugin Creator registration succeeded - Region_TRT
[TRT] Plugin Creator registration succeeded - Clip_TRT
[TRT] Plugin Creator registration succeeded - LReLU_TRT
[TRT] Plugin Creator registration succeeded - PriorBox_TRT
[TRT] Plugin Creator registration succeeded - Normalize_TRT
[TRT] Plugin Creator registration succeeded - RPROI_TRT
[TRT] Plugin Creator registration succeeded - BatchedNMS_TRT
[TRT] completed loading NVIDIA plugins.
[TRT] detected model format - UFF (extension '.uff')
[TRT] desired precision specified for GPU: FASTEST
[TRT] requested fasted precision for device GPU without providing valid calibrator, disabling INT8
[TRT] native precisions detected for GPU: FP32, FP16
[TRT] selecting fastest native precision for GPU: FP16
[TRT] attempting to open engine cache file networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff.1.1.GPU.FP16.engine
[TRT] loading network profile from engine cache... networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff.1.1.GPU.FP16.engine
[TRT] device GPU, networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff loaded
[TRT] Using an engine plan file across different models of devices is not recommended and is likely to affect performance or even cause errors.
[TRT] device GPU, CUDA engine context initialized with 3 bindings
[TRT] binding -- index 0
-- name 'Input'
-- type FP32
-- in/out INPUT
-- # dims 3
-- dim #0 3 (CHANNEL)
-- dim #1 300 (SPATIAL)
-- dim #2 300 (SPATIAL)
[TRT] binding -- index 1
-- name 'NMS'
-- type FP32
-- in/out OUTPUT
-- # dims 3
-- dim #0 1 (CHANNEL)
-- dim #1 100 (SPATIAL)
-- dim #2 7 (SPATIAL)
[TRT] binding -- index 2
-- name 'NMS_1'
-- type FP32
-- in/out OUTPUT
-- # dims 3
-- dim #0 1 (CHANNEL)
-- dim #1 1 (SPATIAL)
-- dim #2 1 (SPATIAL)
[TRT] binding to input 0 Input binding index: 0
[TRT] binding to input 0 Input dims (b=1 c=3 h=300 w=300) size=1080000
[TRT] binding to output 0 NMS binding index: 1
[TRT] binding to output 0 NMS dims (b=1 c=1 h=100 w=7) size=2800
[TRT] binding to output 1 NMS_1 binding index: 2
[TRT] binding to output 1 NMS_1 dims (b=1 c=1 h=1 w=1) size=4
device GPU, networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff initialized.
W = 7 H = 100 C = 1
detectNet -- maximum bounding boxes: 100
detectNet -- loaded 91 class info entries
detectNet -- number of object classes: 91
[image] loaded './images/peds_0.jpg' (1920 x 1080, 3 channels)
4 objects detected
detected obj 0 class #1 (person) confidence=0.984694
bounding box 0 (1073.928711, 108.778038) (1396.850952, 1002.921448) w=322.922241 h=894.143433
detected obj 1 class #1 (person) confidence=0.699254
bounding box 1 (458.434509, 122.990280) (741.858093, 943.342163) w=283.423584 h=820.351868
detected obj 2 class #1 (person) confidence=0.975855
bounding box 2 (742.781067, 160.996078) (1022.465393, 941.796326) w=279.684326 h=780.800232
detected obj 3 class #1 (person) confidence=0.863392
bounding box 3 (1477.790283, 149.502762) (1720.689331, 914.901855) w=242.899048 h=765.399109
[TRT] ------------------------------------------------
[TRT] Timing Report networks/SSD-Mobilenet-v2/ssd_mobilenet_v2_coco.uff
[TRT] ------------------------------------------------
[TRT] Pre-Process CPU 0.09542ms CUDA 0.77667ms
[TRT] Network CPU 117.64365ms CUDA 116.89984ms
[TRT] Post-Process CPU 0.08344ms CUDA 0.08234ms
[TRT] Visualize CPU 16.16084ms CUDA 16.58760ms
[TRT] Total CPU 133.98334ms CUDA 134.34645ms
[TRT] ------------------------------------------------
[TRT] note -- when processing a single image, run 'sudo jetson_clocks' before
to disable DVFS for more accurate profiling/timing measurements
detectnet-console: writing 1920x1080 image to 'output0.jpg'
detectnet-console: successfully wrote 1920x1080 image to 'output0.jpg'
detectnet-console: shutting down...
detectnet-console: shutdown complete
