yolov5 detect.py调试内容记录

最新推荐文章于 2024-09-30 14:52:48 发布

构建的乐趣

最新推荐文章于 2024-09-30 14:52:48 发布

阅读量785

点赞数 1

文章标签： yolov5

本文链接：https://blog.csdn.net/qq_44065334/article/details/115335053

版权

该博客详细记录了如何在Visual Studio Code（VSCode）中利用YOLOv5x预训练模型进行目标检测。通过带参数调试，展示了模型对图像的预测结果，包括边界框坐标、置信度等信息，并进行了预测结果的筛选和分析。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

采用vs code带参数调试，使用yolov5x预训练模型

type(pred)
<class 'list'>
len(pred)
1
pred[0]
tensor([[ 7.62500e+00,  1.36625e+02,  4.77000e+02,  4.38000e+02,  9.34570e-01,  5.00000e+00],
        [ 3.00625e+01,  2.35375e+02,  1.45500e+02,  5.35500e+02,  9.26270e-01,  0.00000e+00],
        [ 3.98250e+02,  2.31750e+02,  4.79750e+02,  5.22000e+02,  9.22363e-01,  0.00000e+00],
        [ 1.31250e+02,  2.42625e+02,  2.03250e+02,  5.10500e+02,  8.99414e-01,  0.00000e+00],
        [-6.25000e-02,  2.87000e+02,  4.71875e+01,  5.17500e+02,  5.84473e-01,  0.00000e+00],
        [ 3.33750e+02,  1.33750e+01,  3.50750e+02,  3.22500e+01,  2.56592e-01,  5.80000e+01]], device='cuda:0')
pred[0].shape
torch.Size([6, 6])
gn
tensor([ 810, 1080,  810, 1080])
s
'640x480 '
im0
array([[[122, 148, 172],
        [120, 146, 170],
        [125, 153, 177],
        ...,
        [157, 170, 184],
        [158, 171, 185],
        [158, 171, 185]],

       [[127, 153, 177],
        [124, 150, 174],
        [127, 155, 179],
        ...,
        [158, 171, 185],
        [159, 172, 186],
        [159, 172, 186]],

       [[128, 154, 178],
        [126, 152, 176],
        [126, 154, 178],
        ...,
        [158, 171, 185],
        [158, 171, 185],
        [158, 171, 185]],

       ...,

       [[185, 185, 191],
        [182, 182, 188],
        [179, 179, 185],
        ...,
        [114, 107, 112],
        [115, 105, 111],
        [116, 106, 112]],

       [[157, 157, 163],
        [180, 180, 186],
        [185, 186, 190],
        ...,
        [107,  97, 103],
        [102,  92,  98],
        [108,  98, 104]],

       [[112, 112, 118],
        [160, 160, 166],
        [169, 170, 174],
        ...,
        [ 99,  89,  95],
        [ 96,  86,  92],
        [102,  92,  98]]], dtype=uint8)
No debugger available, can not send 'variables'
type(im0)
<class 'numpy.ndarray'>
im0.shape
(1080, 810, 3)
c
tensor(0., device='cuda:0')
det[:,-1]
tensor([ 5.,  0.,  0.,  0.,  0., 58.], device='cuda:0')
det[:,-1].shape
torch.Size([6])
det.shape
torch.Size([6, 6])
det[:, -1].unique()
tensor([ 0.,  5., 58.], device='cuda:0')
det[:, -1].unique().shape
torch.Size([3])
n
tensor(1, device='cuda:0')
det[:, -1].unique()
tensor([ 0.,  5., 58.], device='cuda:0')
Unable to find thread to evaluate variable reference.
det[:, -1] == c
tensor([ True, False, False, False, False, False], device='cuda:0')
det
tensor([[1.30000e+01, 2.31000e+02, 8.05000e+02, 7.39000e+02, 9.34570e-01, 5.00000e+00],
        [5.10000e+01, 3.97000e+02, 2.46000e+02, 9.04000e+02, 9.26270e-01, 0.00000e+00],
        [6.72000e+02, 3.91000e+02, 8.10000e+02, 8.81000e+02, 9.22363e-01, 0.00000e+00],
        [2.21000e+02, 4.09000e+02, 3.43000e+02, 8.61000e+02, 8.99414e-01, 0.00000e+00],
        [0.00000e+00, 4.84000e+02, 8.00000e+01, 8.73000e+02, 5.84473e-01, 0.00000e+00],
        [5.63000e+02, 2.30000e+01, 5.92000e+02, 5.40000e+01, 2.56592e-01, 5.80000e+01]], device='cuda:0')
c
tensor(58., device='cuda:0')
c
tensor(58., device='cuda:0')