[Bug] MMOCRInferencer cropped image 0 size which breaks the recognition

[juvebogdan]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• The bug has not been fixed in the latest version (0.x) or latest version (1.x).

Task

I have modified the scripts/configs, or I'm working on my own tasks/models/datasets.

Branch

main branch https://github.com/open-mmlab/mmocr

Environment

sys.platform: linux
Python: 3.10.12 (main, Nov 20 2023, 15:14:05) [GCC 11.4.0]
CUDA available: True
numpy_random_seed: 2147483648
GPU 0: Tesla T4
CUDA_HOME: /usr/local/cuda
NVCC: Cuda compilation tools, release 12.2, V12.2.140
GCC: x86_64-linux-gnu-gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0
PyTorch: 2.0.0+cu117
PyTorch compiling details: PyTorch built with:

• GCC 9.3
• C++ Version: 201703
• Intel(R) oneAPI Math Kernel Library Version 2022.2-Product Build 20220804 for Intel(R) 64 architecture applications
• Intel(R) MKL-DNN v2.7.3 (Git Hash 6dbeffbae1f23cbbeae17adb7b5b13f1f37c080e)
• OpenMP 201511 (a.k.a. OpenMP 4.5)
• LAPACK is enabled (usually provided by MKL)
• NNPACK is enabled
• CPU capability usage: AVX2
• CUDA Runtime 11.7
• NVCC architecture flags: -gencode;arch=compute_37,code=sm_37;-gencode;arch=compute_50,code=sm_50;-gencode;arch=compute_60,code=sm_60;-gencode;arch=compute_70,code=sm_70;-gencode;arch=compute_75,code=sm_75;-gencode;arch=compute_80,code=sm_80;-gencode;arch=compute_86,code=sm_86
• CuDNN 8.5
• Magma 2.6.1
• Build settings: BLAS_INFO=mkl, BUILD_TYPE=Release, CUDA_VERSION=11.7, CUDNN_VERSION=8.5.0, CXX_COMPILER=/opt/rh/devtoolset-9/root/usr/bin/c++, CXX_FLAGS= -D_GLIBCXX_USE_CXX11_ABI=0 -fabi-version=11 -Wno-deprecated -fvisibility-inlines-hidden -DUSE_PTHREADPOOL -DNDEBUG -DUSE_KINETO -DLIBKINETO_NOROCTRACER -DUSE_FBGEMM -DUSE_QNNPACK -DUSE_PYTORCH_QNNPACK -DUSE_XNNPACK -DSYMBOLICATE_MOBILE_DEBUG_HANDLE -O2 -fPIC -Wall -Wextra -Werror=return-type -Werror=non-virtual-dtor -Werror=bool-operation -Wnarrowing -Wno-missing-field-initializers -Wno-type-limits -Wno-array-bounds -Wno-unknown-pragmas -Wunused-local-typedefs -Wno-unused-parameter -Wno-unused-function -Wno-unused-result -Wno-strict-overflow -Wno-strict-aliasing -Wno-error=deprecated-declarations -Wno-stringop-overflow -Wno-psabi -Wno-error=pedantic -Wno-error=redundant-decls -Wno-error=old-style-cast -fdiagnostics-color=always -faligned-new -Wno-unused-but-set-variable -Wno-maybe-uninitialized -fno-math-errno -fno-trapping-math -Werror=format -Werror=cast-function-type -Wno-stringop-overflow, LAPACK_INFO=mkl, PERF_WITH_AVX=1, PERF_WITH_AVX2=1, PERF_WITH_AVX512=1, TORCH_DISABLE_GPU_ASSERTS=ON, TORCH_VERSION=2.0.0, USE_CUDA=ON, USE_CUDNN=ON, USE_EXCEPTION_PTR=1, USE_GFLAGS=OFF, USE_GLOG=OFF, USE_MKL=ON, USE_MKLDNN=ON, USE_MPI=OFF, USE_NCCL=1, USE_NNPACK=ON, USE_OPENMP=ON, USE_ROCM=OFF,

TorchVision: 0.15.1+cu117
OpenCV: 4.8.0
MMEngine: 0.10.2
MMOCR: 1.0.1+

Reproduces the problem - code sample

!pip install torch==2.0.0 torchvision==0.15.1
!pip install -U openmim
!mim install "mmengine>=0.7.1,<1.1.0"
!mim install "mmcv>=2.0.0rc4,<2.1.0"
!mim install "mmdet>=3.0.0rc5,<3.2.0"
!mim install mmocr

!git clone https://github.com/open-mmlab/mmocr.git
%cd mmocr
!pip install -v -e .

from mmocr.apis import MMOCRInferencer
infer = MMOCRInferencer(det='psenet', rec='robustscanner')
result = infer('/content/drit_page-0005.jpg', save_vis=True, return_vis=True)
print(result['predictions'])

Reproduces the problem - command or script

!pip install torch==2.0.0 torchvision==0.15.1
!pip install -U openmim
!mim install "mmengine>=0.7.1,<1.1.0"
!mim install "mmcv>=2.0.0rc4,<2.1.0"
!mim install "mmdet>=3.0.0rc5,<3.2.0"
!mim install mmocr

!git clone https://github.com/open-mmlab/mmocr.git
%cd mmocr
!pip install -v -e .

from mmocr.apis import MMOCRInferencer
infer = MMOCRInferencer(det='psenet', rec='robustscanner')
result = infer('/content/drit_page-0005.jpg', save_vis=True, return_vis=True)
print(result['predictions'])

Reproduces the problem - error message

---

error Traceback (most recent call last)
in <cell line: 3>()
1 from mmocr.apis import MMOCRInferencer
2 infer = MMOCRInferencer(det='psenet', rec='robustscanner')
----> 3 result = infer('/content/drit_page-0005.jpg', save_vis=True, return_vis=True)
4 print(result['predictions'])

15 frames
/usr/local/lib/python3.10/dist-packages/mmcv/image/geometric.py in imresize(img, size, return_scale, interpolation, out, backend)
114 resized_img = np.array(pil_image)
115 else:
--> 116 resized_img = cv2.resize(
117 img, size, dst=out, interpolation=cv2_interp_codes[interpolation])
118 if not return_scale:

error: OpenCV(4.8.0) /io/opencv/modules/imgproc/src/resize.cpp:4062: error: (-215:Assertion failed) !ssize.empty() in function 'resize'

Additional information

When i inference with this image i get the error. I believe there is a problem with detection. When i print crop_img(img, quad) shape i get:
(57, 239, 3)
(57, 198, 3)
(57, 251, 3)
(73, 137, 3)
(49, 75, 3)
(58, 168, 3)
(58, 109, 3)
(81, 267, 3)
(83, 304, 3)
(74, 260, 3)
(66, 221, 3)
(82, 153, 3)
(66, 279, 3)
(58, 139, 3)
(58, 251, 3)
(74, 295, 3)
(58, 150, 3)
(58, 227, 3)
(66, 232, 3)
(66, 144, 3)
(66, 268, 3)
(66, 214, 3)
(66, 239, 3)
(66, 185, 3)
(50, 152, 3)
(41, 58, 3)
(41, 64, 3)
(81, 272, 3)
(66, 173, 3)
(58, 151, 3)
(66, 297, 3)
(82, 259, 3)
(33, 48, 3)
(66, 250, 3)
(58, 222, 3)
(58, 269, 3)
(66, 197, 3)
(66, 221, 3)
(41, 65, 3)
(44, 73, 3)
(66, 273, 3)
(74, 284, 3)
(74, 243, 3)
(66, 262, 3)
(49, 147, 3)
(75, 172, 3)
(66, 344, 3)
(66, 221, 3)
(58, 228, 3)
(58, 168, 3)
(57, 157, 3)
(66, 273, 3)
(57, 209, 3)
(66, 173, 3)
(66, 226, 3)
(49, 158, 3)
(58, 198, 3)
(80, 209, 3)
(82, 372, 3)
(58, 151, 3)
(58, 239, 3)
(58, 216, 3)
(66, 191, 3)
(58, 174, 3)
(33, 60, 3)
(57, 156, 3)
(66, 167, 3)
(66, 209, 3)
(57, 127, 3)
(57, 210, 3)
(66, 197, 3)
(57, 222, 3)
(66, 203, 3)
(80, 249, 3)
(49, 25, 3)
(49, 134, 3)
(57, 216, 3)
(66, 215, 3)
(57, 151, 3)
(66, 273, 3)
(58, 240, 3)
(72, 136, 3)
(50, 75, 3)
(66, 215, 3)
(66, 179, 3)
(58, 115, 3)
(50, 122, 3)
(58, 115, 3)
(58, 151, 3)
(80, 319, 3)
(33, 17, 3)
(95, 213, 3)
(58, 140, 3)
(66, 191, 3)
(58, 204, 3)
(21, 33, 3)
(58, 162, 3)
(58, 145, 3)
(74, 290, 3)
(58, 174, 3)
(66, 126, 3)
(12, 0, 3)
(0, 0, 3)
(65, 132, 3)
there are these with (12,0,3) or (0,0,3) that break the system. I tried to remove them from inference but i didn't manage to.

What can i do to prevent this?