Module opencv::dnn

Deep Neural Network module

This module contains:

• API for new layers creation, layers are building bricks of neural networks;
• set of built-in most-useful Layers;
• API to construct and modify comprehensive neural networks from layers;
• functionality for loading serialized networks models from different frameworks.

Functionality of this module is designed only for forward pass computations (i.e. network testing). A network training is in principle not supported.

Structs

AbsLayer
BNLLLayer
BackendNode	Derivatives of this class encapsulates functions of certain backends.
BatchNormLayer
BlankLayer	Partial List of Implemented Layers
ChannelsPReLULayer
ConcatLayer
ConstLayer	Constant layer produces the same data blob at an every forward pass.
ConvolutionLayer
CropAndResizeLayer
CropLayer
DeconvolutionLayer
DetectionOutputLayer
DictValue	This struct stores the scalar value (or array) of one of the following type: double, cv::String or int64. @todo Maybe int64 is useless because double type exactly stores at least 2^52 integers.
ELULayer
EltwiseLayer
FlattenLayer
InnerProductLayer
InterpLayer	Bilinear resize layer from https://github.com/cdmh/deeplab-public-ver2
LRNLayer
LayerFactory	%Layer factory allows to create instances of registered layers.
LayerParams	This class provides all data needed to initialize layer.
MVNLayer
MaxUnpoolLayer
Net	This class allows to create and manipulate comprehensive artificial neural networks.
NormalizeBBoxLayer	inline formula - normalization layer.
PaddingLayer	Adds extra values for specific axes.
PermuteLayer
PoolingLayer
PowerLayer
PriorBoxLayer
ProposalLayer
ReLU6Layer
ReLULayer
RegionLayer
ReorgLayer
ReshapeLayer
ResizeLayer	Resize input 4-dimensional blob by nearest neighbor or bilinear strategy.
ScaleLayer
ShiftLayer
ShuffleChannelLayer	Permute channels of 4-dimensional input blob.
SigmoidLayer
SliceLayer	Slice layer has several modes:
SoftmaxLayer
SplitLayer
TanHLayer

Enums

Backend	Enum of computation backends supported by layers. @see Net::setPreferableBackend
Target	Enum of target devices for computations. @see Net::setPreferableTarget

Constants

CV_DNN_INFERENCE_ENGINE_VPU_TYPE_MYRIAD_2
CV_DNN_INFERENCE_ENGINE_VPU_TYPE_MYRIAD_X
CV_DNN_INFERENCE_ENGINE_VPU_TYPE_UNSPECIFIED
DNN_BACKEND_DEFAULT
DNN_BACKEND_HALIDE
DNN_BACKEND_INFERENCE_ENGINE	Intel's Inference Engine computational backend.
DNN_BACKEND_OPENCV
DNN_TARGET_CPU
DNN_TARGET_FPGA	FPGA device with CPU fallbacks using Inference Engine's Heterogeneous plugin.
DNN_TARGET_MYRIAD
DNN_TARGET_OPENCL
DNN_TARGET_OPENCL_FP16

Traits

ActivationLayer
BackendWrapper	Derivatives of this class wraps cv::Mat for different backends and targets.
BaseConvolutionLayer
Dict	This class implements name-value dictionary, values are instances of DictValue.
LSTMLayer	LSTM recurrent layer
Layer	This interface class allows to build new Layers - are building blocks of networks.
RNNLayer	Classical recurrent layer

Functions

blob_from_image	Creates 4-dimensional blob from image. Optionally resizes and crops @p image from center, subtract @p mean values, scales values by @p scalefactor, swap Blue and Red channels.
blob_from_image_to	Creates 4-dimensional blob from image. @details This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.
blob_from_images	Creates 4-dimensional blob from series of images. Optionally resizes and crops @p images from center, subtract @p mean values, scales values by @p scalefactor, swap Blue and Red channels.
blob_from_images_to	Creates 4-dimensional blob from series of images. @details This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.
clamp
clamp_range
get_available_targets
get_inference_engine_vpu_type	Returns Inference Engine VPU type.
get_plane
images_from_blob	Parse a 4D blob and output the images it contains as 2D arrays through a simpler data structure (std::vectorcv::Mat).
nms_boxes	Performs non maximum suppression given boxes and corresponding scores.
nms_boxes_f64	C++ default parameters
nms_boxes_rotated	C++ default parameters
read_net	Read deep learning network represented in one of the supported formats.
read_net_1	Read deep learning network represented in one of the supported formats. @details This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.
read_net_from_caffe	Reads a network model stored in Caffe framework's format.
read_net_from_caffe_buffer	Reads a network model stored in Caffe model in memory.
read_net_from_caffe_str	Reads a network model stored in Caffe model in memory. @details This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.
read_net_from_darknet	Reads a network model stored in Darknet model files.
read_net_from_darknet_buffer	Reads a network model stored in Darknet model files.
read_net_from_darknet_str	Reads a network model stored in Darknet model files.
read_net_from_model_optimizer	Load a network from Intel's Model Optimizer intermediate representation.
read_net_from_onnx	Reads a network model ONNX.
read_net_from_onnx_buffer	Reads a network model from ONNX in-memory buffer.
read_net_from_onnx_str	Reads a network model from ONNX in-memory buffer.
read_net_from_tensorflow	Reads a network model stored in TensorFlow framework's format.
read_net_from_tensorflow_buffer	Reads a network model stored in TensorFlow framework's format.
read_net_from_tensorflow_str	Reads a network model stored in TensorFlow framework's format. @details This is an overloaded member function, provided for convenience. It differs from the above function only in what argument(s) it accepts.
read_net_from_torch	Reads a network model stored in Torch7 framework's format.
read_tensor_from_onnx	Creates blob from .pb file.
read_torch_blob	Loads blob which was serialized as torch.Tensor object of Torch7 framework. @warning This function has the same limitations as readNetFromTorch().
reset_myriad_device	Release a Myriad device (binded by OpenCV).
shape
shape_3d	C++ default parameters
shape_nd
shape_umat
shrink_caffe_model	Convert all weights of Caffe network to half precision floating point.
slice_1d
slice_2d
slice_3d
slice_4d
write_text_graph	Create a text representation for a binary network stored in protocol buffer format.