OpenCV3.4的神经网络功能主要提供了以下三种：
-
ml模块中的多层感知机（Artificial Neural Networks - Multi-Layer Perceptrons），提供了MLP的创建、训练、参数设置等函数。如：
static Ptr< ANN_MLP > create () Creates empty model. static Ptr< ANN_MLP > load (const String &filepath) Loads and creates a serialized ANN from a file. void setAnnealFinalT (double val) void setAnnealInitialT (double val) void setAnnealItePerStep (int val) virtual void setBackpropMomentumScale (double val)=0 virtual void setBackpropWeightScale (double val)=0 virtual void setLayerSizes (InputArray _layer_sizes)=0 virtual void setRpropDW0 (double val)=0 virtual void setRpropDWMax (double val)=0 enum ActivationFunctions { IDENTITY = 0, SIGMOID_SYM = 1, GAUSSIAN = 2, RELU = 3, LEAKYRELU = 4 } enum TrainFlags { UPDATE_WEIGHTS = 1, NO_INPUT_SCALE = 2, NO_OUTPUT_SCALE = 4 } enum TrainingMethods { BACKPROP =0, RPROP = 1, ANNEAL = 2 }
请参看帮助文档。
-
DNN模块，提供了很多用于创建、加载、训练深度网络和参数设置以及加载TensorFlow、Caffe、Torch模型的方法和类，如：
class cv::dnn::BackendNode Derivatives of this class encapsulates functions of certain backends. class cv::dnn::BackendWrapper Derivatives of this class wraps cv::Mat for different backends and targets. class cv::dnn::Dict This class implements name-value dictionary, values are instances of DictValue. struct cv::dnn::DictValue This struct stores the scalar value (or array) of one of the following type: double, cv::String or int64. class cv::dnn::Layer This interface class allows to build new Layers - are building blocks of networks. class cv::dnn::LayerParams This class provides all data needed to initialize layer. class cv::dnn::Net This class allows to create and manipulate comprehensive artificial neural networks. Mat cv::dnn::blobFromImages (const std::vector< Mat > &images, double scalefactor=1.0, Size size=Size(), const Scalar &mean=Scalar(), bool swapRB=true, bool crop=true) Creates 4-dimensional blob from series of images. Optionally resizes and crops images from center, subtract mean values, scales values by scalefactor, swap Blue and Red channels. void cv::dnn::NMSBoxes (const std::vector< Rect > &bboxes, const std::vector< float > &scores, const float score_threshold, const float nms_threshold, std::vector< int > &indices, const float eta=1.f, const int top_k=0) Performs non maximum suppression given boxes and corresponding scores. Net cv::dnn::readNetFromCaffe (const String &prototxt, const String &caffeModel=String()) Reads a network model stored in Caffe framework's format. Net cv::dnn::readNetFromDarknet (const String &cfgFile, const String &darknetModel=String()) Reads a network model stored in Darknet model files. Net cv::dnn::readNetFromTensorflow (const String &model, const String &config=String()) Reads a network model stored in TensorFlow framework's format. Net cv::dnn::readNetFromTorch (const String &model, bool isBinary=true)参看帮助文档。
- 第三方深度网络工具，详情请查看帮助文档。
下面给出示例。
1.基于MLP的识别。该程序人工生成四类动物数据，通过MLP网络训练模型并检测测试数据类型。
#exam1.py
import cv2
import numpy as np
from random import randint
#创建MLP网络，并设置训练方法、激活函数、层大小和迭代终止条件。
animals_net = cv2.ml.ANN_MLP_create()
animals_net.setTrainMethod(cv2.ml.ANN_MLP_RPROP | cv2.ml.ANN_MLP_UPDATE_WEIGHTS)
animals_net.setActivationFunction(cv2.ml.ANN_MLP_SIGMOID_SYM)
animals_net.setLayerSizes(np.array([3, 6, 4]))
animals_net.setTermCriteria(( cv2.TERM_CRITERIA_EPS | cv2.TERM_CRITERIA_COUNT, 10, 1 ))
#生成四类动物数据及类标记
def dog_sample():
return [randint(10, 20), 1, randint(38, 42)]
def dog_class():
return [1, 0, 0, 0]
def condor_sample():
return [randint(3,10), randint(3,5), 0]
def condor_class():
return [0, 1, 0, 0]
def dolphin_sample():
return [randint(30, 190), randint(5, 15), randint(80, 100)]
def dolphin_class():
return [0, 0, 1, 0]
def dragon_sample():
return [randint(1200, 1800), randint(30, 40), randint(160, 180)]
def dragon_class():
return [0, 0, 0, 1]
#将动物数据和类标记组成一个记录（样本）
def record(sample, classification):
return (np.array([sample], dtype=np.float32), np.array([classification], dtype=np.float32))
#获取5000个样本数据
records = []
RECORDS = 5000
for x in range(0, RECORDS):
records.append(record(dog_sample(), dog_class()))
records.append(record(condor_sample(), condor_class()))
records.append(record(dolphin_sample(), dolphin_class()))
records.append(record(dragon_sample(), dragon_class()))
#训练MLP网络
EPOCHS = 2
for e in range(0, EPOCHS):
print("Epoch %d:" % e)
for t, c in records:
animals_net.train(t, cv2.ml.ROW_SAMPLE, c)
#预测测试样本类别
TESTS = 100
dog_results = 0
for x in range(0, TESTS):
clas = int(animals_net.predict(np.array([dog_sample()], dtype=np.float32))[0])
print("class: %d" % clas)
if (clas) == 0:
dog_results += 1
condor_results = 0
for x in range(0, TESTS):
clas = int(animals_net.predict(np.array([condor_sample()], dtype=np.float32))[0])
print("class: %d" % clas)
if (clas) == 1:
condor_results += 1
dolphin_results = 0
for x in range(0, TESTS):
clas = int(animals_net.predict(np.array([dolphin_sample()], dtype=np.float32))[0])
print("class: %d" % clas)
if (clas) == 2:
dolphin_results += 1
dragon_results = 0
for x in range(0, TESTS):
clas = int(animals_net.predict(np.array([dragon_sample()], dtype=np.float32))[0])
print("class: %d" % clas)
if (clas) == 3:
dragon_results += 1
#输出测试准确率
print("Dog accuracy: %f%%" % (dog_results))
print("condor accuracy: %f%%" % (condor_results))
print("dolphin accuracy: %f%%" % (dolphin_results))
print("dragon accuracy: %f%%" % (dragon_results))
2.基于DNN的识别。该程序加载预先训练的caffe模型在摄像头获取的图像上检测人脸。
import numpy as np
import argparse
import cv2 as cv
#若出现ImportError，请配置环境变量PYTHONPATH为Python可执行文件的地址。
#若不能解决，请更新相关包（或卸载后重新安装）。
try:
import cv2 as cv
except ImportError:
raise ImportError('Can\'t find OpenCV Python module. If you\'ve built it from sources without installation, '
'configure environemnt variable PYTHONPATH to "opencv_build_dir/lib" directory (with "python3" subdirectory if required)')
#导入DNN模块
from cv2 import dnn
inWidth = 300
inHeight = 300
confThreshold = 0.5
#该文件包含在opencv3.4\sources\samples\dnn\face_detector目录中，该目录的上级目录为OpenCV3.4的下载或安装目录
prototxt = 'face_detector/deploy.prototxt'
#该caffe模型文件需先下载，请参看opencv3.4\sources\samples\dnn\face_detector目录中的文本文件
caffemodel = 'face_detector/res10_300x300_ssd_iter_140000.caffemodel'
#加载caffe模型并从摄像头获取图像
if __name__ == '__main__':
net = dnn.readNetFromCaffe(prototxt, caffemodel)
cap = cv.VideoCapture(0)
while True:
ret, frame = cap.read()
cols = frame.shape[1]
rows = frame.shape[0]
#将获取的图像设置为网络输入，设置网络传播方向，检测人脸
net.setInput(dnn.blobFromImage(frame, 1.0, (inWidth, inHeight), (104.0, 177.0, 123.0), False, False))
detections = net.forward()
perf_stats = net.getPerfProfile()
print('Inference time, ms: %.2f' % (perf_stats[0] / cv.getTickFrequency() * 1000))
for i in range(detections.shape[2]):
confidence = detections[0, 0, i, 2]
if confidence > confThreshold:
xLeftBottom = int(detections[0, 0, i, 3] * cols)
yLeftBottom = int(detections[0, 0, i, 4] * rows)
xRightTop = int(detections[0, 0, i, 5] * cols)
yRightTop = int(detections[0, 0, i, 6] * rows)
cv.rectangle(frame, (xLeftBottom, yLeftBottom), (xRightTop, yRightTop),
(0, 255, 0))
label = "face: %.4f" % confidence
labelSize, baseLine = cv.getTextSize(label, cv.FONT_HERSHEY_SIMPLEX, 0.5, 1)
cv.rectangle(frame, (xLeftBottom, yLeftBottom - labelSize[1]),
(xLeftBottom + labelSize[0], yLeftBottom + baseLine),
(255, 255, 255), cv.FILLED)
cv.putText(frame, label, (xLeftBottom, yLeftBottom),
cv.FONT_HERSHEY_SIMPLEX, 0.5, (0, 0, 0))
cv.imshow("detections", frame)
if cv.waitKey(1) != -1:
break