Import error : undefined symbol: load_image_cv #47
Description
I try to install with GPU + OpenCV + CUDNN + OpenMP (Darknet is built with GPU + OpenCV + CUDNN + OpenMP as well)
install script (it is running in a docker container as root so no need for sudo):
#!/bin/bash
rm -rf build
rm /usr/local/lib/python3.5/dist-packages/pyyolo*
make
python3 setup_gpu.py build
python3 setup_gpu.py install
python3 -c "import pyyolo;"
I get the following after running it:
Traceback (most recent call last):
File "<string>", line 1, in <module>
ImportError: /usr/local/lib/python3.5/dist-packages/pyyolo.cpython-35m-x86_64-linux-gnu.so: undefined symbol: load_image_cv
This is the MakeFile:
GPU=1
CUDNN=1
DEBUG=0
OPENCV=1
# ARCH= -gencode arch=compute_20,code=[sm_20,sm_21] \
# -gencode arch=compute_30,code=sm_30 \
# -gencode arch=compute_35,code=sm_35 \
# -gencode arch=compute_50,code=[sm_50,compute_50] \
# -gencode arch=compute_52,code=[sm_52,compute_52]
# This is what I use, uncomment if you know your arch and want to specify
ARCH= -gencode arch=compute_61,code=compute_61 # 1080 TI
# ARCH= -gencode arch=compute_61,code=compute_61 # Tegra TX2
VPATH=./darknet/src/:./darknet/examples
LIB=libyolo.a
OBJDIR=./obj/
CC=gcc -fopenmp -lm -lmvec -lpthread `pkg-config --libs opencv` -lstdc++
AR=ar
NVCC=nvcc
OPTS=-Ofast
COMMON=
CFLAGS=-Wall -Wfatal-errors -Wno-unused-result -fPIC
CFLAGS+=-I./darknet/src -I./darknet/include
ifeq ($(DEBUG), 1)
OPTS=-O0 -g
endif
CFLAGS+=$(OPTS)
ifeq ($(OPENCV), 1)
COMMON+= -DOPENCV
CFLAGS+= -DOPENCV
COMMON+= `pkg-config --libs --cflags opencv` -lstdc++
endif
ifeq ($(GPU), 1)
COMMON+= -DGPU -I/usr/local/cuda/include/
CFLAGS+= -DGPU
endif
ifeq ($(CUDNN), 1)
COMMON+= -DCUDNN
CFLAGS+= -DCUDNN
endif
OBJ=libyolo.o gemm.o utils.o cuda.o deconvolutional_layer.o convolutional_layer.o list.o image.o activations.o im2col.o col2im.o blas.o crop_layer.o dropout_layer.o maxpool_layer.o softmax_layer.o data.o matrix.o network.o connected_layer.o cost_layer.o parser.o option_list.o
OBJ+=detection_layer.o route_layer.o upsample_layer.o box.o normalization_layer.o avgpool_layer.o layer.o local_layer.o shortcut_layer.o logistic_layer.o activation_layer.o rnn_layer.o gru_layer.o crnn_layer.o demo.o batchnorm_layer.o region_layer.o reorg_layer.o tree.o lstm_layer.o l2norm_layer.o yolo_layer.o
OBJ+=captcha.o lsd.o super.o art.o tag.o cifar.o go.o rnn.o segmenter.o regressor.o classifier.o coco.o yolo.o detector.o nightmare.o darknet.o
ifeq ($(GPU), 1)
OBJ+=convolutional_kernels.o deconvolutional_kernels.o activation_kernels.o im2col_kernels.o col2im_kernels.o blas_kernels.o crop_layer_kernels.o dropout_layer_kernels.o maxpool_layer_kernels.o avgpool_layer_kernels.o # network_kernels.o
endif
OBJS = $(addprefix $(OBJDIR), $(OBJ))
DEPS = $(wildcard src/*.h) Makefile
all: obj $(LIB)
$(LIB): $(OBJS)
$(AR) rcs $@ $^
$(OBJDIR)%.o: %.c $(DEPS)
$(CC) $(COMMON) $(CFLAGS) -c $< -o $@
$(OBJDIR)%.o: %.cu $(DEPS)
$(NVCC) $(ARCH) $(COMMON) --compiler-options "$(CFLAGS)" -c $< -o $@
obj:
mkdir -p obj
.PHONY: clean
clean:
rm -rf $(OBJS) $(LIB)
setup_gpu.py:
from distutils.core import setup, Extension
import numpy
module = Extension('pyyolo',
library_dirs=['.', '/usr/local/cuda/lib64', '/usr/local/'],
libraries=['yolo', 'cuda', 'cudart', 'cublas', 'curand', 'cudnn'],
include_dirs=[numpy.get_include(), './darknet/include'],
sources = ['module.c'],
extra_compile_args=['-fopenmp'],
extra_link_args=['-lgomp', '-lm', '-lmvec', '-lpthread', '-lstdc++'])
setup (name = 'pyyolo',
version = '0.1',
description = 'YOLO wrapper',
ext_modules = [module])
I built OpenCV 3.4.3 from source.
cuDNN v7
Python 3.5
CUDA 9.0
Thank you very much.