【香橙派】使用NPU部署Yolov5的完整解决方案

最新推荐文章于 2024-07-05 04:13:20 发布

宏沉一笑

最新推荐文章于 2024-07-05 04:13:20 发布

阅读量1.9k

点赞数 11

分类专栏： # Python 文章标签： YOLO

本文链接：https://blog.csdn.net/w1002285542/article/details/137108960

版权

Python 专栏收录该内容

1 篇文章 0 订阅

订阅专栏

香橙派-RKNN-Yolov5

环境说明

RKNN版本：1.4.0
香橙派系统环境：Ubuntu20
香橙派python版本：3.7
电脑系统环境：Ubuntu20
电脑python版本：3.8
硬件设备：香橙派5

所需源码

https://github.com/rockchip-linux/rknn-toolkit2.git

https://github.com/rockchip-linux/rknpu2.git

https://github.com/ultralytics/yolov5.git

RKNN 香橙派部署

代码下载

git clone https://github.com/rockchip-linux/rknn-toolkit2.git
git checkout v1.6.0

查看RKNN 1.6.0版本支持的python版本

ls rknn_toolkit_lite2/packages
# 结果 
rknn_toolkit_lite2-1.6.0-cp310-cp310-linux_aarch64.whl  
rknn_toolkit_lite2-1.6.0-cp37-cp37m-linux_aarch64.whl  
rknn_toolkit_lite2-1.6.0-cp39-cp39-linux_aarch64.whl
rknn_toolkit_lite2-1.6.0-cp311-cp311-linux_aarch64.whl  
rknn_toolkit_lite2-1.6.0-cp38-cp38-linux_aarch64.whl   
rknn_toolkit_lite2_1.6.0_packages.md5sum

我们可以使用python 3.7 到 3.11之间的任何版本。

安装python 3.7

conda create -n py37 python=3.7
conda activate py37
pip install rknn_toolkit_lite2/packages/rknn_toolkit_lite2-1.6.0-cp37-cp37m-linux_aarch64.whl
pip install opencv-python

运行测试程序

cd rknn_toolkit_lite2/examples/dynamic_shape
python test.py

运行结果

--> Load RKNN model
done
--> Init runtime environment
I RKNN: [11:16:18.618] RKNN Runtime Information: librknnrt version: 1.4.0 (a10f100eb@2022-09-09T09:07:14)
I RKNN: [11:16:18.618] RKNN Driver Information: version: 0.8.2
E RKNN: [11:16:18.618] 6, 1
E RKNN: [11:16:18.618] Invalid RKNN model version 6
E RKNN: [11:16:18.618] rknn_init, load model failed!
E Catch exception when init runtime!
E Traceback (most recent call last):
  File "/opt/conda/envs/py37/lib/python3.7/site-packages/rknnlite/api/rknn_lite.py", line 148, in init_runtime
    self.rknn_runtime.build_graph(self.rknn_data, self.load_model_in_npu)
  File "rknnlite/api/rknn_runtime.py", line 919, in rknnlite.api.rknn_runtime.RKNNRuntime.build_graph
Exception: RKNN init failed. error code: RKNN_ERR_FAIL

Init runtime environment failed

从运行的结果中librknnrt version: 1.4.0，可以看出。我们的设备支持RKNN的1.4.0的版本
接下来，我们就安装RKNN1.4.0的版本。

RKNN 1.4.0版本安装

切换 RKNN 1.4.0版本

git checkout v1.4.0

支持的Python版本查看

ls rknn_toolkit_lite2/packages
@结果 支持python 3.7与3.9
rknn_toolkit_lite2-1.4.0-cp37-cp37m-linux_aarch64.whl  
rknn_toolkit_lite2-1.4.0-cp39-cp39-linux_aarch64.whl  
rknn_toolkit_lite2_1.4.0_packages.md5sum

安装Python3.7环境

conda create -n py37 python=3.7
conda activate py37

安装rknn-toolkit_lite2

pip install rknn_toolkit_lite2/packages/rknn_toolkit_lite2-1.4.0-cp37-cp37m-linux_aarch64.whl
pip install opencv-python

测试安装效果

# 没有出现报错，则安装成功
python -c "from rknnlite.api import RKNNLite"

RKNN PC Linux部署

由于香橙派上只能用1.4.0的版本，电脑上也用1.4.0的版本

下载指定tag的代码

git clone https://github.com/rockchip-linux/rknn-toolkit2.git
git checkout v1.4.0

查看RKNN 1.6.0版本支持的python版本

ls packages/
# 结果 
md5sum.txt  
rknn_toolkit2-1.4.0_22dcfef4-cp36-cp36m-linux_x86_64.whl  
rknn_toolkit2-1.4.0_22dcfef4-cp38-cp38-linux_x86_64.whl

我们可以使用python 3.6 和 3.8的版本。
3. 创建Python3.8的环境

conda create -n py38 python=3.8
conda activate py38

安装rknn_toolkit2

pip install packages/rknn_toolkit2-1.4.0_22dcfef4-cp38-cp38-linux_x86_64.whl

运行yolov5例程

cd examples/onnx/yolov5
python test.py
# 结果
--> Export rknn model
done
--> Init runtime environment
W init_runtime: Target is None, use simulator!
done
--> Running model
Analysing : 100%|███████████████████████████████████████████████| 146/146 [00:00<00:00, 5137.71it/s]
Preparing : 100%|███████████████████████████████████████████████| 146/146 [00:00<00:00, 1165.36it/s]
W inference: The dims of input(ndarray) shape (640, 640, 3) is wrong, expect dims is 4! Try expand dims to (1, 640, 640, 3)!
done
class: person, score: 0.8223356008529663
box coordinate left,top,right,down: [473.26745200157166, 231.93780636787415, 562.1268351078033, 519.7597033977509]
class: person, score: 0.817978024482727
box coordinate left,top,right,down: [211.9896697998047, 245.0290389060974, 283.70787048339844, 513.9374527931213]
class: person, score: 0.7971192598342896
box coordinate left,top,right,down: [115.24964022636414, 232.44154334068298, 207.7837154865265, 546.1097872257233]
class: person, score: 0.4627230763435364
box coordinate left,top,right,down: [79.09242534637451, 339.18042743206024, 121.60038471221924, 514.234916806221]
class: bus , score: 0.7545359134674072
box coordinate left,top,right,down: [86.41703361272812, 134.41848754882812, 558.1083570122719, 460.4184875488281]

目录下的yolov5s.rknn就是最后生成的模型。
6. 适配3588模型

# 239行  可以不修改
rknn = RKNN(verbose=True)
替换为
rknn = RKNN()

# 243行
rknn.config(mean_values=[128, 128, 128], std_values=[128, 128, 128])
替换为
rknn.config(mean_values=[128, 128, 128], std_values=[128, 128, 128], target_platform='rk3588')

# 272行
ret = rknn.init_runtime()
替换为
ret = rknn.init_runtime(target='rk3588')

运行test.py，生成yolov5s.rknn

python test.py
# 结果
done
--> Export rknn model
done
--> Init runtime environment
I target set by user is: rk3588
I Starting ntp or adb, target is RK3588
I Start adb...
error: no devices/emulators found
E init_runtime: Connect to Device Failure (-1), Please make sure the USB connection is normal!
E init_runtime: Catch exception when init runtime!
E init_runtime: Traceback (most recent call last):
E init_runtime:   File "rknn/api/rknn_base.py", line 1975, in rknn.api.rknn_base.RKNNBase.init_runtime
E init_runtime:   File "rknn/api/rknn_runtime.py", line 194, in rknn.api.rknn_runtime.RKNNRuntime.__init__
E init_runtime:   File "rknn/api/rknn_platform.py", line 331, in rknn.api.rknn_platform.start_ntp_or_adb
E init_runtime: Exception: Init runtime environment failed!
Init runtime environment failed!

将yolov5s.rknn传到香橙派上

香橙派运行yolov5

建立文件deploy.py，将下面的内容写入文件中。

import numpy as np
import cv2
from rknnlite.api import RKNNLite

RKNN_MODEL = 'yolov5s.rknn'
IMG_PATH = "bus.jpg"

QUANTIZE_ON = True

OBJ_THRESH = 0.25
NMS_THRESH = 0.45
IMG_SIZE = 640

CLASSES = ("person", "bicycle", "car", "motorbike ", "aeroplane ", "bus ", "train", "truck ", "boat", "traffic light",
           "fire hydrant", "stop sign ", "parking meter", "bench", "bird", "cat", "dog ", "horse ", "sheep", "cow", "elephant",
           "bear", "zebra ", "giraffe", "backpack", "umbrella", "handbag", "tie", "suitcase", "frisbee", "skis", "snowboard", "sports ball", "kite",
           "baseball bat", "baseball glove", "skateboard", "surfboard", "tennis racket", "bottle", "wine glass", "cup", "fork", "knife ",
           "spoon", "bowl", "banana", "apple", "sandwich", "orange", "broccoli", "carrot", "hot dog", "pizza ", "donut", "cake", "chair", "sofa",
           "pottedplant", "bed", "diningtable", "toilet ", "tvmonitor", "laptop	", "mouse	", "remote ", "keyboard ", "cell phone", "microwave ",
           "oven ", "toaster", "sink", "refrigerator ", "book", "clock", "vase", "scissors ", "teddy bear ", "hair drier", "toothbrush ")


def sigmoid(x):
    return 1 / (1 + np.exp(-x))


def xywh2xyxy(x):
    # Convert [x, y, w, h] to [x1, y1, x2, y2]
    y = np.copy(x)
    y[:, 0] = x[:, 0] - x[:, 2] / 2  # top left x
    y[:, 1] = x[:, 1] - x[:, 3] / 2  # top left y
    y[:, 2] = x[:, 0] + x[:, 2] / 2  # bottom right x
    y[:, 3] = x[:, 1] + x[:, 3] / 2  # bottom right y
    return y


def process(input, mask, anchors):

    anchors = [anchors[i] for i in mask]
    grid_h, grid_w = map(int, input.shape[0:2])

    box_confidence = sigmoid(input[..., 4])
    box_confidence = np.expand_dims(box_confidence, axis=-1)

    box_class_probs = sigmoid(input[..., 5:])

    box_xy = sigmoid(input[..., :2])*2 - 0.5

    col = np.tile(np.arange(0, grid_w), grid_w).reshape(-1, grid_w)
    row = np.tile(np.arange(0, grid_h).reshape(-1, 1), grid_h)
    col = col.reshape(grid_h, grid_w, 1, 1).repeat(3, axis=-2)
    row = row.reshape(grid_h, grid_w, 1, 1).repeat(3, axis=-2)
    grid = np.concatenate((col, row), axis=-1)
    box_xy += grid
    box_xy *= int(IMG_SIZE/grid_h)

    box_wh = pow(sigmoid(input[..., 2:4])*2, 2)
    box_wh = box_wh * anchors

    box = np.concatenate((box_xy, box_wh), axis=-1)

    return box, box_confidence, box_class_probs


def filter_boxes(boxes, box_confidences, box_class_probs):
    """Filter boxes with box threshold. It's a bit different with origin yolov5 post process!

    # Arguments
        boxes: ndarray, boxes of objects.
        box_confidences: ndarray, confidences of objects.
        box_class_probs: ndarray, class_probs of objects.

    # Returns
        boxes: ndarray, filtered boxes.
        classes: ndarray, classes for boxes.
        scores: ndarray, scores for boxes.
    """
    boxes = boxes.reshape(-1, 4)
    box_confidences = box_confidences.reshape(-1)
    box_class_probs = box_class_probs.reshape(-1, box_class_probs.shape[-1])

    _box_pos = np.where(box_confidences >= OBJ_THRESH)
    boxes = boxes[_box_pos]
    box_confidences = box_confidences[_box_pos]
    box_class_probs = box_class_probs[_box_pos]

    class_max_score = np.max(box_class_probs, axis=-1)
    classes = np.argmax(box_class_probs, axis=-1)
    _class_pos = np.where(class_max_score >= OBJ_THRESH)

    boxes = boxes[_class_pos]
    classes = classes[_class_pos]
    scores = (class_max_score* box_confidences)[_class_pos]

    return boxes, classes, scores


def nms_boxes(boxes, scores):
    """Suppress non-maximal boxes.

    # Arguments
        boxes: ndarray, boxes of objects.
        scores: ndarray, scores of objects.

    # Returns
        keep: ndarray, index of effective boxes.
    """
    x = boxes[:, 0]
    y = boxes[:, 1]
    w = boxes[:, 2] - boxes[:, 0]
    h = boxes[:, 3] - boxes[:, 1]

    areas = w * h
    order = scores.argsort()[::-1]

    keep = []
    while order.size > 0:
        i = order[0]
        keep.append(i)

        xx1 = np.maximum(x[i], x[order[1:]])
        yy1 = np.maximum(y[i], y[order[1:]])
        xx2 = np.minimum(x[i] + w[i], x[order[1:]] + w[order[1:]])
        yy2 = np.minimum(y[i] + h[i], y[order[1:]] + h[order[1:]])

        w1 = np.maximum(0.0, xx2 - xx1 + 0.00001)
        h1 = np.maximum(0.0, yy2 - yy1 + 0.00001)
        inter = w1 * h1

        ovr = inter / (areas[i] + areas[order[1:]] - inter)
        inds = np.where(ovr <= NMS_THRESH)[0]
        order = order[inds + 1]
    keep = np.array(keep)
    return keep


def yolov5_post_process(input_data):
    masks = [[0, 1, 2], [3, 4, 5], [6, 7, 8]]
    anchors = [[10, 13], [16, 30], [33, 23], [30, 61], [62, 45],
               [59, 119], [116, 90], [156, 198], [373, 326]]

    boxes, classes, scores = [], [], []
    for input, mask in zip(input_data, masks):
        b, c, s = process(input, mask, anchors)
        b, c, s = filter_boxes(b, c, s)
        boxes.append(b)
        classes.append(c)
        scores.append(s)

    boxes = np.concatenate(boxes)
    boxes = xywh2xyxy(boxes)
    classes = np.concatenate(classes)
    scores = np.concatenate(scores)

    nboxes, nclasses, nscores = [], [], []
    for c in set(classes):
        inds = np.where(classes == c)
        b = boxes[inds]
        c = classes[inds]
        s = scores[inds]

        keep = nms_boxes(b, s)

        nboxes.append(b[keep])
        nclasses.append(c[keep])
        nscores.append(s[keep])

    if not nclasses and not nscores:
        return None, None, None

    boxes = np.concatenate(nboxes)
    classes = np.concatenate(nclasses)
    scores = np.concatenate(nscores)

    return boxes, classes, scores


def draw(image, boxes, scores, classes):
    """Draw the boxes on the image.

    # Argument:
        image: original image.
        boxes: ndarray, boxes of objects.
        classes: ndarray, classes of objects.
        scores: ndarray, scores of objects.
        all_classes: all classes name.
    """
    for box, score, cl in zip(boxes, scores, classes):
        top, left, right, bottom = box
        print('class: {}, score: {}'.format(CLASSES[cl], score))
        print('box coordinate left,top,right,down: [{}, {}, {}, {}]'.format(top, left, right, bottom))
        top = int(top)
        left = int(left)
        right = int(right)
        bottom = int(bottom)

        cv2.rectangle(image, (top, left), (right, bottom), (255, 0, 0), 2)
        cv2.putText(image, '{0} {1:.2f}'.format(CLASSES[cl], score),
                    (top, left - 6),
                    cv2.FONT_HERSHEY_SIMPLEX,
                    0.6, (0, 0, 255), 2)


def letterbox(im, new_shape=(640, 640), color=(0, 0, 0)):
    # Resize and pad image while meeting stride-multiple constraints
    shape = im.shape[:2]  # current shape [height, width]
    if isinstance(new_shape, int):
        new_shape = (new_shape, new_shape)

    # Scale ratio (new / old)
    r = min(new_shape[0] / shape[0], new_shape[1] / shape[1])

    # Compute padding
    ratio = r, r  # width, height ratios
    new_unpad = int(round(shape[1] * r)), int(round(shape[0] * r))
    dw, dh = new_shape[1] - new_unpad[0], new_shape[0] - new_unpad[1]  # wh padding

    dw /= 2  # divide padding into 2 sides
    dh /= 2

    if shape[::-1] != new_unpad:  # resize
        im = cv2.resize(im, new_unpad, interpolation=cv2.INTER_LINEAR)
    top, bottom = int(round(dh - 0.1)), int(round(dh + 0.1))
    left, right = int(round(dw - 0.1)), int(round(dw + 0.1))
    im = cv2.copyMakeBorder(im, top, bottom, left, right, cv2.BORDER_CONSTANT, value=color)  # add border
    return im, ratio, (dw, dh)


if __name__ == '__main__':

    # Create RKNN object
    rknn = RKNNLite()

    # load RKNN model
    print('--> Load RKNN model')
    ret = rknn.load_rknn(RKNN_MODEL)


    # Init runtime environment
    print('--> Init runtime environment')
    ret = rknn.init_runtime(core_mask=RKNNLite.NPU_CORE_0_1_2)  #使用0 1 2三个NPU核心
    # ret = rknn.init_runtime('rk3566')
    if ret != 0:
        print('Init runtime environment failed!')
        exit(ret)
    print('done')

    # Set inputs
    img = cv2.imread(IMG_PATH)
    # img, ratio, (dw, dh) = letterbox(img, new_shape=(IMG_SIZE, IMG_SIZE))
    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
    img = cv2.resize(img, (IMG_SIZE, IMG_SIZE))

    # Inference
    outputs = rknn.inference(inputs=[img])


    # post process
    input0_data = outputs[0]
    input1_data = outputs[1]
    input2_data = outputs[2]

    input0_data = input0_data.reshape([3, -1]+list(input0_data.shape[-2:]))
    input1_data = input1_data.reshape([3, -1]+list(input1_data.shape[-2:]))
    input2_data = input2_data.reshape([3, -1]+list(input2_data.shape[-2:]))

    input_data = list()
    input_data.append(np.transpose(input0_data, (2, 3, 0, 1)))
    input_data.append(np.transpose(input1_data, (2, 3, 0, 1)))
    input_data.append(np.transpose(input2_data, (2, 3, 0, 1)))

    boxes, classes, scores = yolov5_post_process(input_data)

    img_1 = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
    if boxes is not None:
        draw(img_1, boxes, scores, classes)
    # show output
    cv2.imshow("post process result", img_1)
    cv2.waitKey(0)
    cv2.destroyAllWindows()

    rknn.release()

运行程序

python deploy.py

结果展示

--> Load RKNN model
--> Init runtime environment
I RKNN: [13:00:34.125] RKNN Runtime Information: librknnrt version: 1.4.0 (a10f100eb@2022-09-09T09:07:14)
I RKNN: [13:00:34.125] RKNN Driver Information: version: 0.8.2
I RKNN: [13:00:34.125] RKNN Model Information: version: 1, toolkit version: 1.4.0-22dcfef4(compiler version: 1.4.0 (3b4520e4f@2022-09-05T20:52:35)), target: RKNPU v2, target platform: rk3588, framework name: ONNX, framework layout: NCHW
done
class: person, score: 0.8928694725036621
box coordinate left,top,right,down: [40.627504616975784, 235.50569766759872, 193.95716640353203, 534.8807769417763]
class: person, score: 0.878767728805542
box coordinate left,top,right,down: [531.5798395127058, 224.8939728140831, 639.2570745497942, 524.2690520882607]
class: person, score: 0.8309085965156555
box coordinate left,top,right,down: [174.62115895748138, 241.39454793930054, 273.0503498315811, 510.30296182632446]
class: person, score: 0.4691559970378876
box coordinate left,top,right,down: [0.81878662109375, 325.4185321331024, 54.147705078125, 517.9757549762726]
class: bicycle, score: 0.24984873831272125
box coordinate left,top,right,down: [530.9477665424347, 11.197928547859192, 597.0226924419403, 57.32749259471893]
class: bus , score: 0.5245066285133362
box coordinate left,top,right,down: [9.820066332817078, 131.40282291173935, 624.8392354249954, 430.7779021859169]

在这里插入图片描述

帮助文档

源码
https://github.com/rockchip-linux/rknpu2
https://github.com/rockchip-linux/rknn-toolkit2
https://github.com/ultralytics/yolov5
ROC-RK3588-PC
https://wiki.t-firefly.com/zh_CN/ROC-RK3588-PC/usage_npu.html
【边缘设备】yolov5训练与rknn模型导出并在RK3588部署~1.环境准备（亲测有效）
https://blog.csdn.net/zhoujinwang/article/details/132320665
香橙派5 RK3588 yolov5模型转换rknn及部署踩坑全记录 orangepi 5
https://blog.csdn.net/m0_55217834/article/details/130583886