MediaPipe window环境下编译指南（2）

前面的文章已经把Mediapipe的环境配置好了，已经可以编译成apk文件了

接下来我们介绍下如何供android使用的aar文件

在编译aar之前我先介绍下mediapipe提供了哪些可以编译的aar工程

上面的这个截图对应的目录是\rootfs\home\tian\mediapipe\mediapipe\examples\android\src\java\com\google\mediapipe\apps

basic:基础工程，下面的其他工程对它有依赖

具体什么含义，可以自己英文查询下

我用编译facedetectiongpu为例子

1、在\rootfs\home\tian\mediapipe\mediapipe\examples\android\src\java\com\google\mediapipe\apps

下创建aar_example文件

扫描二维码关注公众号，回复： 13395173 查看本文章

然后在aar_example，这里面新建文件BUILD.txt,然后把txt去掉，把下面内容复制进去

load("//mediapipe/java/com/google/mediapipe:mediapipe_aar.bzl", "mediapipe_aar")

mediapipe_aar(
    name = "mp_face_detection_aar",
    calculators = ["//mediapipe/graphs/face_detection:mobile_calculators"],
)

注意：

load("//mediapipe/java/com/google/mediapipe:mediapipe_aar.bzl", "mediapipe_aar")

是固定的，这个是编译aar的固定编译配置

name:这个是随意的，就是你编译完成后aar的名称

calculators ：这个是具体编译aar用到的具体的graphs图资源，路径在/mediapipe/graphs/下面

需要什么，就配置什么，

我们打开看下face_detection中的BUILD看下

# Copyright 2019 The MediaPipe Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#      http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

licenses(["notice"])

package(default_visibility = ["//visibility:public"])

cc_library(
    name = "mobile_calculators",
    deps = [
        "//mediapipe/calculators/core:flow_limiter_calculator",
        "//mediapipe/calculators/util:annotation_overlay_calculator",
        "//mediapipe/calculators/util:detections_to_render_data_calculator",
        "//mediapipe/gpu:gpu_buffer_to_image_frame_calculator",
        "//mediapipe/gpu:image_frame_to_gpu_buffer_calculator",
        "//mediapipe/modules/face_detection:face_detection_front_cpu",
        "//mediapipe/modules/face_detection:face_detection_front_gpu",
    ],
)

cc_library(
    name = "desktop_live_calculators",
    deps = [
        "//mediapipe/calculators/core:flow_limiter_calculator",
        "//mediapipe/calculators/util:annotation_overlay_calculator",
        "//mediapipe/calculators/util:detections_to_render_data_calculator",
        "//mediapipe/modules/face_detection:face_detection_front_cpu",
    ],
)

cc_library(
    name = "desktop_live_gpu_calculators",
    deps = [
        "//mediapipe/calculators/core:flow_limiter_calculator",
        "//mediapipe/calculators/util:annotation_overlay_calculator",
        "//mediapipe/calculators/util:detections_to_render_data_calculator",
        "//mediapipe/modules/face_detection:face_detection_front_gpu",
    ],
)

load(
    "//mediapipe/framework/tool:mediapipe_graph.bzl",
    "mediapipe_binary_graph",
)

mediapipe_binary_graph(
    name = "face_detection_mobile_cpu_binary_graph",
    graph = "face_detection_mobile_cpu.pbtxt",
    output_name = "face_detection_mobile_cpu.binarypb",
    deps = [":mobile_calculators"],
)

mediapipe_binary_graph(
    name = "face_detection_mobile_gpu_binary_graph",
    graph = "face_detection_mobile_gpu.pbtxt",
    output_name = "face_detection_mobile_gpu.binarypb",
    deps = [":mobile_calculators"],
)

看到这个我们怎么选择具体的什么name?

这个在官方文档上没有写，我是直接看官方的app里的build是怎么写看下

apps/facedetectiongpu/BUILD

# Copyright 2019 The MediaPipe Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#      http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

licenses(["notice"])

package(default_visibility = ["//visibility:private"])

cc_binary(
    name = "libmediapipe_jni.so",
    linkshared = 1,
    linkstatic = 1,
    deps = [
        "//mediapipe/graphs/face_detection:mobile_calculators",
        "//mediapipe/java/com/google/mediapipe/framework/jni:mediapipe_framework_jni",
    ],
)

cc_library(
    name = "mediapipe_jni_lib",
    srcs = [":libmediapipe_jni.so"],
    alwayslink = 1,
)

android_binary(
    name = "facedetectiongpu",
    srcs = glob(["*.java"]),
    assets = [
        "//mediapipe/graphs/face_detection:face_detection_mobile_gpu.binarypb",
        "//mediapipe/modules/face_detection:face_detection_front.tflite",
    ],
    assets_dir = "",
    manifest = "//mediapipe/examples/android/src/java/com/google/mediapipe/apps/basic:AndroidManifest.xml",
    manifest_values = {
        "applicationId": "com.google.mediapipe.apps.facedetectiongpu",
        "appName": "Face Detection",
        "mainActivity": "com.google.mediapipe.apps.basic.MainActivity",
        "cameraFacingFront": "True",
        "binaryGraphName": "face_detection_mobile_gpu.binarypb",
        "inputVideoStreamName": "input_video",
        "outputVideoStreamName": "output_video",
        "flipFramesVertically": "True",
        "converterNumBuffers": "2",
    },
    multidex = "native",
    deps = [
        ":mediapipe_jni_lib",
        "//mediapipe/examples/android/src/java/com/google/mediapipe/apps/basic:basic_lib",
    ],
)

cc_binary(
name = "libmediapipe_jni.so",
linkshared = 1,
linkstatic = 1,
deps = [
"//mediapipe/graphs/face_detection:mobile_calculators",
"//mediapipe/java/com/google/mediapipe/framework/jni:mediapipe_framework_jni",
],
)

有上面那个mobile_calculators，这个是就要编译的具体的图配置，

接下来我们开始执行cmd命令

//给aar_example赋读写权限

1、sudo chmod -R 755 /mediapipe/examples/android/src/java/com/google/mediapipe/apps/aar_example //后面就是具体的aar_example 目录了，根据自己情况下

//下面这个指令就是编译aar文件了，会等很久，估计20分钟-40分钟，根据自己电脑配置决定

2、bazel build -c opt --fat_apk_cpu=arm64-v8a,armeabi-v7a //mediapipe/examples/android/src/java/com/google/mediapipe/apps/aar_example:mp_face_detection_aar

这个中间可能会出现很多的TimeOut的错误，因为编译过程会下载很多国外的资源，所以很慢或者超时了，这个就自己看有没有梯子或者切换个网络试试，我也是弄了很久，才好的

编译过程漫长，耐心等待，遇到问题仔细分析，一般都会编译成功

如果看到这个就代表成功了

去什么地方找这个aar呢，我之前用官方的方式没有找到

cp bazel-bin/mediapipe/examples/android/src/java/com/google/mediapipe/apps/aar_example/mp_face_detection_aar.aar
/absolute/path/to/your/preferred/location
这个是官方介绍的，但是bazel-bin/mediapipe/examples/android/src/java/com/google/mediapipe/apps/aar_example/mp_face_detection_aar.aar 
这个目录根本没有，执行这个命令会包，找不到文件的错误

最好我发现在这里

C:\Users\Mr.tian\AppData\Local\Packages\CanonicalGroupLimited.Ubuntu18.04onWindows_79rhkp1fndgsc\LocalState\rootfs\home\tian\.cache\bazel\_bazel_tian\7fe30ee7bfd1145e7a3fb5f9113d3a01\execroot\mediapipe\bazel-out\k8-opt\bin\mediapipe\examples\android\src\java\com\google\mediapipe\apps\

这个根据自己的目录在找，或者直接在\rootfs\home\tian搜索你要的aar文件

祝你好运~

MediaPipe window环境下编译指南（2）

猜你喜欢