三、openCV学习笔记（Image Processing）
（一） Image Filtering
使用Filter时注意边界可能需要插值，borderInterpolate()。
BaseColumnFilter: filters with single-column kernels.
BaseFilters:2D image filters using 2D kernels.
BaseRowFilter:filters with single-row kernels.
FilterEngine:apply an arbitrary filtering operation to an image.
bilateralFilter:
adaptiveBilateralFilter:
blur: Blurs an image using the normalized box filter.
broderInterpolate: Determines the origin position of a pixel that is extrapolated.
boxFilter: Blurs an image using the box filter.
buildPyramid:Constructs the Gaussian pyramid for an image.
copyMakeBroder: Forms a border around an image.
createBoxFilter:
createDerivFilter: Returns an engine for computing image derivatives.
createGaussianFilter: Produces a component capable of blurring images using the Gaussian blur technique.
createLinearFilter:
createMorphologyFilter:
createSeparableFilter:
dilate:
erode:
filter2D: Convolves an image with the kernel.
GaussianBlur:
getDerivKernels:
getGaussianKernels:
getKernelType:
getStructuringElement:
medianBlur:
morphologyEx:Performs advanced morphological transformations.
Laplacian:
pyrDown: Blurs an image and downsamples it.
pyrUp: Upsamples an image and then blurs it.
pyrMeanShiftFiltering:
sepFilter2D:
Smooth:
Sobel:
Scharr:
详细介绍：http://docs.opencv.org/2.4.9/modules/imgproc/doc/filtering.html
（二）Geometric Image Transformations
注意：Extrapolation of non-existing pixels.
Interpolation of pixel values.
包括一些空间变换。
深入解析http://docs.opencv.org/2.4.9/modules/imgproc/doc/geometric_transformations.html#undistortpoints
（三）Miscellaneous Image Transformations
adaptive threshold: Applies an adaptive threshold to an array.
cvtColor: Converts an image from one color space to another.
distanceTransform: Computes the minimal distance to a zero pixel across every pixel in the source image.
floodFill: Fills a connected component with the given color.
intergral: Calculates the integral of an image.
threshold:
This method carries out a marker-driven image partitioning employing the watershed algorithm.
grabCut: 不知道干啥的。
详细说明了http://docs.opencv.org/2.4.9/modules/imgproc/doc/miscellaneous_transformations.html#这一资源。
（四）Histogram
详细介绍：http://docs.opencv.org/2.4.9/modules/imgproc/doc/histograms.html#
（五）Structural Analysis and Shape descriptors
moment: ???
findContours:
该函数通过给定精度逼近多边形曲线
ApproxChains: Approximates Freeman chain(s) with a polygonal curve.
arcLength: Calculates a contour perimeter or a curve length.
boundingRect: Calculates the up-right bounding rectangle of a point set.
coutourArea: Calculates a contour area.
convexHull: Finds the convex hull of a point set.
convexityDefects: Finds the convexity defects of a contour.
fitEllipse:
fitLine:
isContourConvext: Tests a contour convexity.
minAreaRect: Computes the smallest area rotated bounding rectangle for the given 2D point set.
minEnclosingCircle: Computes the smallest enclosing circle that encloses a given set of 2D points.
matchShapes: Compares two shapes.
pointPolygonTest: Performs a point-in-contour test.
详细说明该模块中的内容
（六）Motion analysis and Object tracking
详细介绍：http://docs.opencv.org/2.4.9/modules/imgproc/doc/motion_analysis_and_object_tracking.html#
（七）Feature Detection
Canny
cornerEigenValsAndVecs: Computes the eigenvalues and eigenvectors of image blocks to identify corners in images.
cornerHarris:Harris edge detector.
cornerMinEigenVal: Computes the minimum eigenvalue of gradient matrices in corner detection.
cornerSubPix: Refines the corner locations.
goodFeaturesToTrack: Determines strong corners on an image.
HoughCircles: Detects circles within grayscale images by employing the Hough transform.
HoughLines identifies straight lines within a binary matrix by employing the classical Hough transform.
HoughLinesP: Detects straight lines in a binary image based on the probabilistic Hough transform.
preCornerDetect: Calculates a feature map for corner detection.
完整展示相关内容的位置如下：http://docs.opencv.org/2.4.9/modules/imgproc/doc/feature_detection.html#
（八）Object Detection
matchTemplate: Compares a template against overlapped image regions.