OpenCV的2.4.7.版本生成了python的CV2模塊，可以直接載入；

有興趣的可以參考這個教程：http://blog.csdn.net/sunny2038/article/details/9080047? 不過可惜的是，這個教程只是針對CV2 的；

Tips1： 關于兩種載入方法的區別：

import numpy as np import cv2 as cv2 Vs. import cv2.cv as cv

（一）：引入CV2使用：

第一種方法的cv函數使用方法幾乎等同于matlab，其中可以使用的函數范圍也較少；

例如：利用help(cv2)命令；

查看函數列表如下：

ANN_MLP([layerSizes[, activateFunc[, fparam1[, fparam2]]]]) -> <ANN_MLP object>Algorithm__create(name) -> retvalAlgorithm_getList() -> algorithmsBFMatcher([, normType[, crossCheck]]) -> <BFMatcher object>BRISK([, thresh[, octaves[, patternScale]]]) -> <BRISK object> or BRISK(radiusList, numberList[, dMax[, dMin[, indexChange]]]) -> <BRISK object>BackgroundSubtractorMOG([history, nmixtures, backgroundRatio[, noiseSigma]]) -> <BackgroundSubtractorMOG object>Boost([trainData, tflag, responses[, varIdx[, sampleIdx[, varType[, missingDataMask[, params]]]]]]) -> <Boost object>CamShift(probImage, window, criteria) -> retval, windowCanny(image, threshold1, threshold2[, edges[, apertureSize[, L2gradient]]]) -> edgesCascadeClassifier([filename]) -> <CascadeClassifier object>DMatch() -> <DMatch object> or DMatch(_queryIdx, _trainIdx, _distance) -> <DMatch object> or DMatch(_queryIdx, _trainIdx, _imgIdx, _distance) -> <DMatch object>DTree() -> <DTree object> DescriptorExtractor_create(descriptorExtractorType) -> retvalDescriptorMatcher_create(descriptorMatcherType) -> retvalEM([, nclusters[, covMatType[, termCrit]]]) -> <EM object>ERTrees() -> <ERTrees object>FastFeatureDetector([, threshold[, nonmaxSuppression]]) -> <FastFeatureDetector object> Feature2D_create(name) -> retvalFeatureDetector_create(detectorType) -> retvalFileNode() -> <FileNode object> FileStorage([source, flags[, encoding]]) -> <FileStorage object>FlannBasedMatcher([, indexParams[, searchParams]]) -> <FlannBasedMatcher object>GBTrees([trainData, tflag, responses[, varIdx[, sampleIdx[, varType[, missingDataMask[, params]]]]]]) -> <GBTrees object>GFTTDetector([, maxCorners[, qualityLevel[, minDistance[, blockSize[, useHarrisDetector[, k]]]]]]) -> <GFTTDetector object>GaussianBlur(src, ksize, sigmaX[, dst[, sigmaY[, borderType]]]) -> dstGridAdaptedFeatureDetector([, detector[, maxTotalKeypoints[, gridRows[, gridCols]]]]) -> <GridAdaptedFeatureDetector object>HOGDescriptor() -> <HOGDescriptor object> or HOGDescriptor(_winSize, _blockSize, _blockStride, _cellSize, _nbins[, _derivAperture[, _winSigma[, _histogramNormType[, _L2HysThreshold[, _gammaCorrection[, _nlevels]]]]]]) -> <HOGDescriptor object> or HOGDescriptor(filename) -> <HOGDescriptor object>HOGDescriptor_getDaimlerPeopleDetector() -> retvalHOGDescriptor_getDefaultPeopleDetector() -> retvalHoughCircles(image, method, dp, minDist[, circles[, param1[, param2[, minRadius[, maxRadius]]]]]) -> circlesHoughLines(image, rho, theta, threshold[, lines[, srn[, stn]]]) -> linesHoughLinesP(image, rho, theta, threshold[, lines[, minLineLength[, maxLineGap]]]) -> linesHuMoments(m[, hu]) -> huKDTree() -> <KDTree object> or KDTree(points[, copyAndReorderPoints]) -> <KDTree object> or KDTree(points, _labels[, copyAndReorderPoints]) -> <KDTree object> KNearest([trainData, responses[, sampleIdx[, isRegression[, max_k]]]]) -> <KNearest object> KalmanFilter([dynamParams, measureParams[, controlParams[, type]]]) -> <KalmanFilter object> KeyPoint([x, y, _size[, _angle[, _response[, _octave[, _class_id]]]]]) -> <KeyPoint object> LUT(src, lut[, dst[, interpolation]]) -> dst Laplacian(src, ddepth[, dst[, ksize[, scale[, delta[, borderType]]]]]) -> dstMSER([, _delta[, _min_area[, _max_area[, _max_variation[, _min_diversity[, _max_evolution[, _area_threshold[, _min_margin[, _edge_blur_size]]]]]]]]]) -> <MSER object>Mahalanobis(v1, v2, icovar) -> retvalNormalBayesClassifier([trainData, responses[, varIdx[, sampleIdx]]]) -> <NormalBayesClassifier object>ORB([, nfeatures[, scaleFactor[, nlevels[, edgeThreshold[, firstLevel[, WTA_K[, scoreType[, patchSize]]]]]]]]) -> <ORB object>PCABackProject(data, mean, eigenvectors[, result]) -> resultPCACompute(data[, mean[, eigenvectors[, maxComponents]]]) -> mean, eigenvectorsPCAComputeVar(data, retainedVariance[, mean[, eigenvectors]]) -> mean, eigenvectorsPCAProject(data, mean, eigenvectors[, result]) -> resultPSNR(src1, src2) -> retvalPyramidAdaptedFeatureDetector(detector[, maxLevel]) -> <PyramidAdaptedFeatureDetector object>RQDecomp3x3(src[, mtxR[, mtxQ[, Qx[, Qy[, Qz]]]]]) -> retval, mtxR, mtxQ, Qx, Qy, QzRTrees() -> <RTrees object>Rodrigues(src[, dst[, jacobian]]) -> dst, jacobianSIFT([, nfeatures[, nOctaveLayers[, contrastThreshold[, edgeThreshold[, sigma]]]]]) -> <SIFT object> SURF([hessianThreshold[, nOctaves[, nOctaveLayers[, extended[, upright]]]]]) -> <SURF object>SVBackSubst(w, u, vt, rhs[, dst]) -> dstSVDecomp(src[, w[, u[, vt[, flags]]]]) -> w, u, vt SVM([trainData, responses[, varIdx[, sampleIdx[, params]]]]) -> <SVM object>Scharr(src, ddepth, dx, dy[, dst[, scale[, delta[, borderType]]]]) -> dstSimpleBlobDetector([, parameters]) -> <SimpleBlobDetector object> SimpleBlobDetector_Params() -> <SimpleBlobDetector_Params object>Sobel(src, ddepth, dx, dy[, dst[, ksize[, scale[, delta[, borderType]]]]]) -> dstStarDetector([, _maxSize[, _responseThreshold[, _lineThresholdProjected[, _lineThresholdBinarized[, _suppressNonmaxSize]]]]]) -> <StarDetector object>StereoBM([preset[, ndisparities[, SADWindowSize]]]) -> <StereoBM object> StereoSGBM([minDisparity, numDisparities, SADWindowSize[, P1[, P2[, disp12MaxDiff[, preFilterCap[, uniquenessRatio[, speckleWindowSize[, speckleRange[, fullDP]]]]]]]]]) -> <StereoSGBM object> StereoVar([levels, pyrScale, nIt, minDisp, maxDisp, poly_n, poly_sigma, fi, lambda, penalization, cycle, flags]) -> <StereoVar object>Subdiv2D([rect]) -> <Subdiv2D object>VideoCapture() -> <VideoCapture object> or VideoCapture(filename) -> <VideoCapture object> or VideoCapture(device) -> <VideoCapture object>VideoWriter([filename, fourcc, fps, frameSize[, isColor]]) -> <VideoWriter object>absdiff(src1, src2[, dst]) -> dst accumulate(src, dst[, mask]) -> None accumulateProduct(src1, src2, dst[, mask]) -> None accumulateSquare(src, dst[, mask]) -> None accumulateWeighted(src, dst, alpha[, mask]) -> None adaptiveBilateralFilter(src, ksize, sigmaSpace[, dst[, maxSigmaColor[, anchor[, borderType]]]]) -> dst adaptiveThreshold(src, maxValue, adaptiveMethod, thresholdType, blockSize, C[, dst]) -> dst add(src1, src2[, dst[, mask[, dtype]]]) -> dst addWeighted(src1, alpha, src2, beta, gamma[, dst[, dtype]]) -> dst applyColorMap(src, colormap[, dst]) -> dstapproxPolyDP(curve, epsilon, closed[, approxCurve]) -> approxCurve arcLength(curve, closed) -> retval batchDistance(src1, src2, dtype[, dist[, nidx[, normType[, K[, mask[, update[, crosscheck]]]]]]]) -> dist, nidx bilateralFilter(src, d, sigmaColor, sigmaSpace[, dst[, borderType]]) -> dst bitwise_and(src1, src2[, dst[, mask]]) -> dst bitwise_not(src[, dst[, mask]]) -> dst bitwise_or(src1, src2[, dst[, mask]]) -> dst bitwise_xor(src1, src2[, dst[, mask]]) -> dstblur(src, ksize[, dst[, anchor[, borderType]]]) -> dst borderInterpolate(p, len, borderType) -> retval boundingRect(points) -> retval boxFilter(src, ddepth, ksize[, dst[, anchor[, normalize[, borderType]]]]) -> dst buildOpticalFlowPyramid(img, winSize, maxLevel[, pyramid[, withDerivatives[, pyrBorder[, derivBorder[, tryReuseInputImage]]]]]) -> retval, pyramidcalcBackProject(images, channels, hist, ranges, scale[, dst]) -> dst calcCovarMatrix(samples, flags[, covar[, mean[, ctype]]]) -> covar, mean calcGlobalOrientation(orientation, mask, mhi, timestamp, duration) -> retval calcHist(images, channels, mask, histSize, ranges[, hist[, accumulate]]) -> hist calcMotionGradient(mhi, delta1, delta2[, mask[, orientation[, apertureSize]]]) -> mask, orientation calcOpticalFlowFarneback(prev, next, pyr_scale, levels, winsize, iterations, poly_n, poly_sigma, flags[, flow]) -> flow calcOpticalFlowPyrLK(prevImg, nextImg, prevPts[, nextPts[, status[, err[, winSize[, maxLevel[, criteria[, flags[, minEigThreshold]]]]]]]]) -> nextPts, status, err calcOpticalFlowSF(from, to, flow, layers, averaging_block_size, max_flow) -> None or calcOpticalFlowSF(from, to, flow, layers, averaging_block_size, max_flow,sigma_dist, sigma_color, postprocess_window, sigma_dist_fix, sigma_color_fix, occ_thr, upscale_averaging_radius, upscale_sigma_dist, upscale_sigma_color, speed_up_thr) -> NonecalibrateCamera(objectPoints, imagePoints, imageSize[, cameraMatrix[, distCoeffs[, rvecs[, tvecs[, flags[, criteria]]]]]]) -> retval, cameraMatrix, distCoeffs, rvecs, tvecs calibrationMatrixValues(cameraMatrix, imageSize, apertureWidth, apertureHeight) -> fovx, fovy, focalLength, principalPoint, aspectRatio cartToPolar(x, y[, magnitude[, angle[, angleInDegrees]]]) -> magnitude, anglechamerMatching(img, templ[, templScale[, maxMatches[, minMatchDistance[, padX[, padY[, scales[, minScale[, maxScale[, orientationWeight[, truncate]]]]]]]]]]) -> retval, results, costcheckHardwareSupport(feature) -> retval checkRange(a[, quiet[, minVal[, maxVal]]]) -> retval, pos circle(img, center, radius, color[, thickness[, lineType[, shift]]]) -> None clipLine(imgRect, pt1, pt2) -> retval, pt1, pt2 compare(src1, src2, cmpop[, dst]) -> dst compareHist(H1, H2, method) -> retval completeSymm(mtx[, lowerToUpper]) -> None composeRT(rvec1, tvec1, rvec2, tvec2[, rvec3[, tvec3[, dr3dr1[, dr3dt1[, dr3dr2[, dr3dt2[, dt3dr1[, dt3dt1[, dt3dr2[, dt3dt2]]]]]]]]]]) -> rvec3, tvec3, dr3dr1, dr3dt1, dr3dr2, dr3dt2, dt3dr1, dt3dt1, dt3dr2, dt3dt2computeCorrespondEpilines(points, whichImage, F[, lines]) -> lines contourArea(contour[, oriented]) -> retval convertMaps(map1, map2, dstmap1type[, dstmap1[, dstmap2[, nninterpolation]]]) -> dstmap1, dstmap2 convertPointsFromHomogeneous(src[, dst]) -> dst convertPointsToHomogeneous(src[, dst]) -> dst convertScaleAbs(src[, dst[, alpha[, beta]]]) -> dstconvexHull(points[, hull[, clockwise[, returnPoints]]]) -> hull convexityDefects(contour, convexhull[, convexityDefects]) -> convexityDefects copyMakeBorder(src, top, bottom, left, right, borderType[, dst[, value]]) -> dst cornerEigenValsAndVecs(src, blockSize, ksize[, dst[, borderType]]) -> dst cornerHarris(src, blockSize, ksize, k[, dst[, borderType]]) -> dst cornerMinEigenVal(src, blockSize[, dst[, ksize[, borderType]]]) -> dstcornerSubPix(image, corners, winSize, zeroZone, criteria) -> None correctMatches(F, points1, points2[, newPoints1[, newPoints2]]) -> newPoints1, newPoints2 countNonZero(src) -> retvalcreateCLAHE([, clipLimit[, tileGridSize]]) -> retval createEigenFaceRecognizer([, num_components[, threshold]]) -> retval createFisherFaceRecognizer([, num_components[, threshold]]) -> retval createHanningWindow(winSize, type[, dst]) -> dst createLBPHFaceRecognizer([, radius[, neighbors[, grid_x[, grid_y[, threshold]]]]]) -> retvalcreateTrackbar(trackbarName, windowName, value, count, onChange) -> None cubeRoot(val) -> retval cvtColor(src, code[, dst[, dstCn]]) -> dst dct(src[, dst[, flags]]) -> dstdecomposeProjectionMatrix(projMatrix[, cameraMatrix[, rotMatrix[, transVect[, rotMatrixX[, rotMatrixY[, rotMatrixZ[, eulerAngles]]]]]]]) -> cameraMatrix, rotMatrix, transVect, rotMatrixX, rotMatrixY, rotMatrixZ, eulerAnglesdestroyAllWindows() -> None destroyWindow(winname) -> None determinant(mtx) -> retvaldft(src[, dst[, flags[, nonzeroRows]]]) -> dst dilate(src, kernel[, dst[, anchor[, iterations[, borderType[, borderValue]]]]]) -> dst distanceTransform(src, distanceType, maskSize[, dst]) -> dst distanceTransformWithLabels(src, distanceType, maskSize[, dst[, labels[, labelType]]]) -> dst, labels divide(src1, src2[, dst[, scale[, dtype]]]) -> dst or divide(scale, src2[, dst[, dtype]]) -> dst drawChessboardCorners(image, patternSize, corners, patternWasFound) -> None drawContours(image, contours, contourIdx, color[, thickness[, lineType[, hierarchy[, maxLevel[, offset]]]]]) -> NonedrawDataMatrixCodes(image, codes, corners) -> None drawKeypoints(image, keypoints[, outImage[, color[, flags]]]) -> outImage eigen(src, computeEigenvectors[, eigenvalues[, eigenvectors]]) -> retval, eigenvalues, eigenvectors ellipse(img, center, axes, angle, startAngle, endAngle, color[, thickness[, lineType[, shift]]]) -> None or ellipse(img, box, color[, thickness[, lineType]]) -> None ellipse2Poly(center, axes, angle, arcStart, arcEnd, delta) -> pts equalizeHist(src[, dst]) -> dsterode(src, kernel[, dst[, anchor[, iterations[, borderType[, borderValue]]]]]) -> dst estimateAffine3D(src, dst[, out[, inliers[, ransacThreshold[, confidence]]]]) -> retval, out, inliers estimateRigidTransform(src, dst, fullAffine) -> retval exp(src[, dst]) -> dst extractChannel(src, coi[, dst]) -> dst fastAtan2(y, x) -> retvalfastNlMeansDenoising(src[, dst[, h[, templateWindowSize[, searchWindowSize]]]]) -> dst fastNlMeansDenoisingColored(src[, dst[, h[, hColor[, templateWindowSize[, searchWindowSize]]]]]) -> dst fastNlMeansDenoisingColoredMulti(srcImgs, imgToDenoiseIndex, temporalWindowSize[, dst[, h[, hColor[, templateWindowSize[, searchWindowSize]]]]]) -> dst fastNlMeansDenoisingMulti(srcImgs, imgToDenoiseIndex, temporalWindowSize[, dst[, h[, templateWindowSize[, searchWindowSize]]]]) -> dst fillConvexPoly(img, points, color[, lineType[, shift]]) -> None fillPoly(img, pts, color[, lineType[, shift[, offset]]]) -> Nonefilter2D(src, ddepth, kernel[, dst[, anchor[, delta[, borderType]]]]) -> dst filterSpeckles(img, newVal, maxSpeckleSize, maxDiff[, buf]) -> None findChessboardCorners(image, patternSize[, corners[, flags]]) -> retval, corners findCirclesGrid(image, patternSize[, centers[, flags[, blobDetector]]]) -> retval, centers findCirclesGridDefault(image, patternSize[, centers[, flags]]) -> retval, centers findContours(image, mode, method[, contours[, hierarchy[, offset]]]) -> contours, hierarchyfindDataMatrix(image[, corners[, dmtx]]) -> codes, corners, dmtx findFundamentalMat(points1, points2[, method[, param1[, param2[, mask]]]]) -> retval, mask findHomography(srcPoints, dstPoints[, method[, ransacReprojThreshold[, mask]]]) -> retval, mask findNonZero(src[, idx]) -> idx fitEllipse(points) -> retval fitLine(points, distType, param, reps, aeps[, line]) -> lineflann_Index([features, params[, distType]]) -> <flann_Index object> flip(src, flipCode[, dst]) -> dst floodFill(image, mask, seedPoint, newVal[, loDiff[, upDiff[, flags]]]) -> retval, rect gemm(src1, src2, alpha, src3, gamma[, dst[, flags]]) -> dst getAffineTransform(src, dst) -> retval getBuildInformation() -> retval getCPUTickCount() -> retval getDefaultNewCameraMatrix(cameraMatrix[, imgsize[, centerPrincipalPoint]]) -> retvalgetDerivKernels(dx, dy, ksize[, kx[, ky[, normalize[, ktype]]]]) -> kx, ky getGaborKernel(ksize, sigma, theta, lambd, gamma[, psi[, ktype]]) -> retval getGaussianKernel(ksize, sigma[, ktype]) -> retval getNumberOfCPUs() -> retval getOptimalDFTSize(vecsize) -> retval getOptimalNewCameraMatrix(cameraMatrix, distCoeffs, imageSize, alpha[, newImgSize[, centerPrincipalPoint]]) -> retval, validPixROI getPerspectiveTransform(src, dst) -> retval getRectSubPix(image, patchSize, center[, patch[, patchType]]) -> patchgetRotationMatrix2D(center, angle, scale) -> retval getStructuringElement(shape, ksize[, anchor]) -> retval getTextSize(text, fontFace, fontScale, thickness) -> retval, baseLine getTickCount() -> retval getTickFrequency() -> retval getTrackbarPos(trackbarname, winname) -> retval getValidDisparityROI(roi1, roi2, minDisparity, numberOfDisparities, SADWindowSize) -> retval getWindowProperty(winname, prop_id) -> retvalgoodFeaturesToTrack(image, maxCorners, qualityLevel, minDistance[, corners[, mask[, blockSize[, useHarrisDetector[, k]]]]]) -> corners grabCut(img, mask, rect, bgdModel, fgdModel, iterCount[, mode]) -> None groupRectangles(rectList, groupThreshold[, eps]) -> rectList, weights hconcat(src[, dst]) -> dstidct(src[, dst[, flags]]) -> dst idft(src[, dst[, flags[, nonzeroRows]]]) -> dst imdecode(buf, flags) -> retval imencode(ext, img[, params]) -> retval, bufimread(filename[, flags]) -> retval imshow(winname, mat) -> None imwrite(filename, img[, params]) -> retval inRange(src, lowerb, upperb[, dst]) -> dst initCameraMatrix2D(objectPoints, imagePoints, imageSize[, aspectRatio]) -> retval initModule_nonfree() -> retval initUndistortRectifyMap(cameraMatrix, distCoeffs, R, newCameraMatrix, size, m1type[, map1[, map2]]) -> map1, map2 initWideAngleProjMap(cameraMatrix, distCoeffs, imageSize, destImageWidth, m1type[, map1[, map2[, projType[, alpha]]]]) -> retval, map1, map2inpaint(src, inpaintMask, inpaintRadius, flags[, dst]) -> dst insertChannel(src, dst, coi) -> None integral(src[, sum[, sdepth]]) -> sum integral2(src[, sum[, sqsum[, sdepth]]]) -> sum, sqsum integral3(src[, sum[, sqsum[, tilted[, sdepth]]]]) -> sum, sqsum, tiltedintersectConvexConvex(_p1, _p2[, _p12[, handleNested]]) -> retval, _p12 invert(src[, dst[, flags]]) -> retval, dst invertAffineTransform(M[, iM]) -> iM isContourConvex(contour) -> retvalkmeans(data, K, criteria, attempts, flags[, bestLabels[, centers]]) -> retval, bestLabels, centers line(img, pt1, pt2, color[, thickness[, lineType[, shift]]]) -> None log(src[, dst]) -> dst magnitude(x, y[, magnitude]) -> magnitude matMulDeriv(A, B[, dABdA[, dABdB]]) -> dABdA, dABdBmatchShapes(contour1, contour2, method, parameter) -> retval matchTemplate(image, templ, method[, result]) -> result max(src1, src2[, dst]) -> dst mean(src[, mask]) -> retval meanShift(probImage, window, criteria) -> retval, window meanStdDev(src[, mean[, stddev[, mask]]]) -> mean, stddev medianBlur(src, ksize[, dst]) -> dstmerge(mv[, dst]) -> dst min(src1, src2[, dst]) -> dst minAreaRect(points) -> retval minEnclosingCircle(points) -> center, radius minMaxLoc(src[, mask]) -> minVal, maxVal, minLoc, maxLoc mixChannels(src, dst, fromTo) -> Nonemoments(array[, binaryImage]) -> retval morphologyEx(src, op, kernel[, dst[, anchor[, iterations[, borderType[, borderValue]]]]]) -> dst moveWindow(winname, x, y) -> None mulSpectrums(a, b, flags[, c[, conjB]]) -> c mulTransposed(src, aTa[, dst[, delta[, scale[, dtype]]]]) -> dst multiply(src1, src2[, dst[, scale[, dtype]]]) -> dst namedWindow(winname[, flags]) -> Nonenorm(src1[, normType[, mask]]) -> retval or norm(src1, src2[, normType[, mask]]) -> retval normalize(src[, dst[, alpha[, beta[, norm_type[, dtype[, mask]]]]]]) -> dst patchNaNs(a[, val]) -> None perspectiveTransform(src, m[, dst]) -> dst phase(x, y[, angle[, angleInDegrees]]) -> angle phaseCorrelate(src1, src2[, window]) -> retval phaseCorrelateRes(src1, src2, window) -> retval, responsepointPolygonTest(contour, pt, measureDist) -> retval polarToCart(magnitude, angle[, x[, y[, angleInDegrees]]]) -> x, y polylines(img, pts, isClosed, color[, thickness[, lineType[, shift]]]) -> None pow(src, power[, dst]) -> dst preCornerDetect(src, ksize[, dst[, borderType]]) -> dst projectPoints(objectPoints, rvec, tvec, cameraMatrix, distCoeffs[, imagePoints[, jacobian[, aspectRatio]]]) -> imagePoints, jacobianputText(img, text, org, fontFace, fontScale, color[, thickness[, lineType[, bottomLeftOrigin]]]) -> None pyrDown(src[, dst[, dstsize[, borderType]]]) -> dst pyrMeanShiftFiltering(src, sp, sr[, dst[, maxLevel[, termcrit]]]) -> dst pyrUp(src[, dst[, dstsize[, borderType]]]) -> dst randShuffle(dst[, iterFactor]) -> None randn(dst, mean, stddev) -> None randu(dst, low, high) -> None rectangle(img, pt1, pt2, color[, thickness[, lineType[, shift]]]) -> Nonerectify3Collinear(cameraMatrix1, distCoeffs1, cameraMatrix2, distCoeffs2, cameraMatrix3, distCoeffs3, imgpt1, imgpt3, imageSize, R12, T12, R13, T13, alpha, newImgSize, flags[, R1[, R2[, R3[, P1[, P2[, P3[, Q]]]]]]]) -> retval, R1, R2, R3, P1, P2, P3, Q, roi1, roi2 reduce(src, dim, rtype[, dst[, dtype]]) -> dst remap(src, map1, map2, interpolation[, dst[, borderMode[, borderValue]]]) -> dst repeat(src, ny, nx[, dst]) -> dst reprojectImageTo3D(disparity, Q[, _3dImage[, handleMissingValues[, ddepth]]]) -> _3dImageresize(src, dsize[, dst[, fx[, fy[, interpolation]]]]) -> dst resizeWindow(winname, width, height) -> None scaleAdd(src1, alpha, src2[, dst]) -> dst segmentMotion(mhi, timestamp, segThresh[, segmask]) -> segmask, boundingRects sepFilter2D(src, ddepth, kernelX, kernelY[, dst[, anchor[, delta[, borderType]]]]) -> dst setIdentity(mtx[, s]) -> None setMouseCallback(windowName, onMouse [, param]) -> None setTrackbarPos(trackbarname, winname, pos) -> None setUseOptimized(onoff) -> None setWindowProperty(winname, prop_id, prop_value) -> Nonesolve(src1, src2[, dst[, flags]]) -> retval, dst solveCubic(coeffs[, roots]) -> retval, roots solvePnP(objectPoints, imagePoints, cameraMatrix, distCoeffs[, rvec[, tvec[, useExtrinsicGuess[, flags]]]]) -> retval, rvec, tvecsolvePnPRansac(objectPoints, imagePoints, cameraMatrix, distCoeffs[, rvec[, tvec[, useExtrinsicGuess[, iterationsCount[, reprojectionError[, minInliersCount[, inliers[, flags]]]]]]]]) -> rvec, tvec, inlierssolvePoly(coeffs[, roots[, maxIters]]) -> retval, roots sort(src, flags[, dst]) -> dst sortIdx(src, flags[, dst]) -> dst split(m[, mv]) -> mv sqrt(src[, dst]) -> dst startWindowThread() -> retvalstereoCalibrate(objectPoints, imagePoints1, imagePoints2, imageSize[, cameraMatrix1[, distCoeffs1[, cameraMatrix2[, distCoeffs2[, R[, T[, E[, F[, criteria[, flags]]]]]]]]]]) -> retval, cameraMatrix1, distCoeffs1, cameraMatrix2, distCoeffs2, R, T, E, FstereoRectify(cameraMatrix1, distCoeffs1, cameraMatrix2, distCoeffs2, imageSize, R, T[, R1[, R2[, P1[, P2[, Q[, flags[, alpha[, newImageSize]]]]]]]]) -> R1, R2, P1, P2, Q, validPixROI1, validPixROI2stereoRectifyUncalibrated(points1, points2, F, imgSize[, H1[, H2[, threshold]]]) -> retval, H1, H2 subtract(src1, src2[, dst[, mask[, dtype]]]) -> dst sumElems(src) -> retval threshold(src, thresh, maxval, type[, dst]) -> retval, dst trace(mtx) -> retval transform(src, m[, dst]) -> dsttranspose(src[, dst]) -> dst triangulatePoints(projMatr1, projMatr2, projPoints1, projPoints2[, points4D]) -> points4D undistort(src, cameraMatrix, distCoeffs[, dst[, newCameraMatrix]]) -> dst undistortPoints(src, cameraMatrix, distCoeffs[, dst[, R[, P]]]) -> dst updateMotionHistory(silhouette, mhi, timestamp, duration) -> NoneuseOptimized() -> retval validateDisparity(disparity, cost, minDisparity, numberOfDisparities[, disp12MaxDisp]) -> None vconcat(src[, dst]) -> dstwaitKey([, delay]) -> retval warpAffine(src, M, dsize[, dst[, flags[, borderMode[, borderValue]]]]) -> dst warpPerspective(src, M, dsize[, dst[, flags[, borderMode[, borderValue]]]]) -> dst watershed(image, markers) -> None

總結：

對于CV2的使用，其主要類型是mat型，可以直接下標索引：如 Image [idxY] [idxX] = 65552

? 其中有一個疑惑的問題是：

對于resize的使用，? Image? =cv.resize(image,(w,h))??? 此句函數的使用會 導致：自動把Image 轉化為3通道；這直接讓我沒有辦法使用cv2了。

（二）：引入Cv的使用方法：

引入cv2.cv 后，使用cv便可以找到幾乎在cv2refman里面的所有函數了，此處不再列出，舉個例子說明。

path ="D:/Develope/EclipseWorks/SLICSeg/Recog/ViewX_2.094395 ViewY_0.000000 ViewZ_6.141592.pcd_comb.pcd_label.txtNo_SLIC.png" Image = cv.LoadImageM(path,2) # 原始圖像載入，參數為1 會轉化為 三通道，8bit cv.ShowImage("name", Image) cv.WaitKey(0) ImageHeight= Image.height #Image.shape[0] #Iplimage 沒有shape屬性！ ImageWidth = Image.widthcv.Size = (Image.width, Image.height) cv.Size = cv.GetSize(Image)ImagePatch = cv.CreateImage(cv.Size,16, 1) cv.Copy(Image, ImagePatch)#copy大小必須一致！ cv.SetImageROI(ImagePatch,(0,0,Image.width/2,Image.height/2)) path ="Patch2/"+ "path_"+ “76”+"_" + "S_posX" +"_Image.png" cv.SaveImage(path,ImagePatch) cv.ShowImage("name", ImagePatch) cv.WaitKey(0)ImagePatch = cv.CreateImage((320,240), 16, 1) cv.Resize(Image, ImagePatch, cv.CV_INTER_LINEAR) rows = Image.height cols = Image.width Mat = cv.CreateMat(rows, cols, cv.CV_16UC1) cv.Convert(Image, Mat) Image[24,25] =0 #對于單通道圖像 ImagePatch = cv.CreateImage(cv.Size,8, 3) ImagePatch[24,25][2] =255#對于三通道圖像
至此，基本語法講解告一段落！

總結

以上是生活随笔為你收集整理的Python中OpenCV2. VS. CV1的全部內容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網站內容還不錯，歡迎將生活随笔推薦給好友。