「TypeError: Expected Ptr<cv::UMat> for argument 'src'」について

Question

発生している問題・エラーメッセージ

trainer.run()をすると以下のようなエラーが発生します。

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-19-041e2033e90a> in <module>
----> 1 trainer.run()

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/training/trainer.py in run(self, show_loop_exception_msg)
    374                         f.write('Traceback (most recent call last):\n')
    375                         traceback.print_tb(sys.exc_info()[2])
--> 376             six.reraise(*exc_info)
    377         finally:
    378             for _, entry in extensions:

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/six.py in reraise(tp, value, tb)
    701             if value.__traceback__ is not tb:
    702                 raise value.with_traceback(tb)
--> 703             raise value
    704         finally:
    705             value = None

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/training/trainer.py in run(self, show_loop_exception_msg)
    341                 self.observation = {}
    342                 with reporter.scope(self.observation):
--> 343                     update()
    344                     for name, entry in extensions:
    345                         if entry.trigger(self):

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/training/updaters/standard_updater.py in update(self)
    238 
    239         """
--> 240         self.update_core()
    241         self.iteration += 1
    242 

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/training/updaters/standard_updater.py in update_core(self)
    243     def update_core(self):
    244         iterator = self._iterators['main']
--> 245         batch = iterator.next()
    246         in_arrays = convert._call_converter(
    247             self.converter, batch, self.input_device)

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/iterators/serial_iterator.py in __next__(self)
     75             raise StopIteration
     76 
---> 77         batch = [self.dataset[index] for index in indices]
     78         return batch
     79 

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/iterators/serial_iterator.py in <listcomp>(.0)
     75             raise StopIteration
     76 
---> 77         batch = [self.dataset[index] for index in indices]
     78         return batch
     79 

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/dataset/dataset_mixin.py in __getitem__(self, index)
     65             return [self.get_example(i) for i in index]
     66         else:
---> 67             return self.get_example(index)
     68 
     69     def __len__(self):

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainer/datasets/transform_dataset.py in get_example(self, i)
     50     def get_example(self, i):
     51         in_data = self._dataset[i]
---> 52         return self._transform(in_data)

<ipython-input-11-f268809d3c76> in __call__(self, in_data)
      7         img, bbox, label = in_data
      8         _, H, W = img.shape
----> 9         img = self.faster_rcnn.prepare(img)
     10         _, o_H, o_W = img.shape
     11         scale = o_H / H

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainercv/links/model/faster_rcnn/faster_rcnn.py in prepare(self, img)
    215             scale = self.max_size / max(H, W)
    216 
--> 217         img = resize(img, (int(H * scale), int(W * scale)))
    218 
    219         img = (img - self.mean).astype(np.float32, copy=False)

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainercv/transforms/image/resize.py in resize(img, size, interpolation)
     69     if chainer.config.cv_resize_backend is None:
     70         if _cv2_available:
---> 71             return _resize_cv2(img, size, interpolation)
     72         else:
     73             return _resize_pil(img, size, interpolation)

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainercv/transforms/image/resize.py in _resize_cv2(img, size, interpolation)
     22         cv_interpolation = cv2.INTER_LANCZOS4
     23     H, W = size
---> 24     img = cv2.resize(img, dsize=(W, H), interpolation=cv_interpolation)
     25 
     26     # If input is a grayscale image, cv2 returns a two-dimentional array.

TypeError: Expected Ptr<cv::UMat> for argument 'src'

該当のソースコード

以下のコードはChainerCVのexampleを参考にしております。

import matplotlib.pyplot as plt
%matplotlib inline

import os
import numpy as np

import chainer
import random

from chainercv.chainer_experimental.datasets.sliceable import TupleDataset
from chainercv.links import FasterRCNNVGG16
from chainercv.links.model.faster_rcnn import FasterRCNNTrainChain
from chainer.datasets import TransformDataset
from chainercv import transforms
from chainer import training
from chainer.training import extensions

import SLATS_256 as bbox

data_dir = 'ShinjukuGyoen/SLATS-256'
classes_file = 'ShinjukuGyoen/classes.txt'

file_img_set = os.path.join(data_dir, 'images.npy')
file_bbox_set = os.path.join(data_dir, 'bounding_box_data.npy')
file_object_ids = os.path.join(data_dir, 'object_ids.npy')

file_classes = os.path.join(classes_file, 'classes.txt')

imgs = np.load(file_img_set, allow_pickle=True)
bboxs = np.load(file_bbox_set, allow_pickle=True)
objectIDs = np.load(file_object_ids, allow_pickle=True)

classes = list()
with open(classes_file) as fd:
    for one_line in fd.readlines():
        cl = one_line.split('\n')[0]
        classes.append(cl)
print(classes)

dataset = TupleDataset(('img', imgs), ('bbox', bboxs), ('label', objectIDs))

N = len(dataset)
N_train = (int)(N*0.8)
N_test = N - N_train
print('total:{}, train:{}, test:{}'.format(N, N_train, N_test))

faster_rcnn = FasterRCNNVGG16(n_fg_class=len(classes), 
                              pretrained_model='imagenet')

model = FasterRCNNTrainChain(faster_rcnn)
model.to_cpu()

optimizer = chainer.optimizers.MomentumSGD(lr=0.001, momentum=0.9)
optimizer.setup(model)
optimizer.add_hook(chainer.optimizer_hooks.WeightDecay(rate=0.0005))

class Transform(object):

    def __init__(self, faster_rcnn):
        self.faster_rcnn = faster_rcnn

    def __call__(self, in_data):
        img, bbox, label = in_data
        _, H, W = img.shape
        img = self.faster_rcnn.prepare(img)
        _, o_H, o_W = img.shape
        scale = o_H / H
        bbox = transforms.resize_bbox(bbox, (H, W), (o_H, o_W))

        img, params = transforms.random_flip(
            img, x_random=True, return_param=True)
        bbox = transforms.flip_bbox(
            bbox, (o_H, o_W), x_flip=params['x_flip'])

        return img, bbox, label, scale

idxs = list(np.arange(N))
random.shuffle(idxs)
train_idxs = idxs[:N_train]
test_idxs = idxs[N_train:]

train_data = TransformDataset(dataset[train_idxs], Transform(faster_rcnn))
train_iter = chainer.iterators.SerialIterator(
    train_data, batch_size=1)
test_iter = chainer.iterators.SerialIterator(
    dataset[test_idxs], batch_size=1, repeat=False, shuffle=False)

updater = chainer.training.updaters.StandardUpdater(
    train_iter, optimizer)

n_epoch = 20
out_dir = './out'
trainer = training.Trainer(
    updater, (n_epoch, 'epoch'), out=out_dir)

step_size = 100 

trainer.extend(
    extensions.snapshot_object(model.faster_rcnn, 'snapshot_model.npz'),
    trigger=(n_epoch, 'epoch')) 
trainer.extend(extensions.ExponentialShift('lr', 0.1),
               trigger=(step_size, 'iteration'))
 
log_interval = 1, 'epoch'
plot_interval = 1, 'epoch'
print_interval = 1, 'epoch'

trainer.extend(chainer.training.extensions.observe_lr(),
               trigger=log_interval)

trainer.extend(extensions.LogReport(trigger=log_interval)) 

trainer.extend(extensions.PrintReport(
    ['iteration', 'epoch', 'elapsed_time', 'lr',
     'main/loss',
     'main/roi_loc_loss',
     'main/roi_cls_loss',
     'main/rpn_loc_loss',
     'main/rpn_cls_loss',
     'validation/main/map',
    ]), trigger=print_interval)

trainer.extend(
    extensions.PlotReport(
        ['main/loss'],
        file_name='loss.png', trigger=plot_interval
    ),
    trigger=plot_interval
) 

trainer.extend(extensions.dump_graph('main/loss'))

trainer.run()

試したこと

以下の記事を参考にしてみましたが、うまく行きませんでした。

• [OpenCV TypeError: Expected cv::UMat for argument 'src' - What is this?

](https://stackoverflow.com/questions/54249728/opencv-typeerror-expected-cvumat-for-argument-src-what-is-this)

Answer 1

~/GraduationResearch/GraduationResearch/lib/python3.6/site-packages/chainercv/transforms/image/resize.py in _resize_cv2(img, size, interpolation)
22 cv_interpolation = cv2.INTER_LANCZOS4
23 H, W = size
---> 24 img = cv2.resize(img, dsize=(W, H), interpolation=cv_interpolation)
25
26 # If input is a grayscale image, cv2 returns a two-dimentional array.

上のエラー文を参考に修正したところ解決できました。画像に問題がありました。