%加载MAT文件与地面真实数据。加载=负载(“caltech_cordova1_laneAndVehicleGroundTruth.mat”）;

%创建一个VideoReader对象来读取视频帧。videoName = loaded.videoName;fileReader = VideoReader (videoName);地面真实数据被组织在一个时间表中。gtdata = loaded.groundTruthData;%显示地面真实数据的前几行。头(gtdata)

阅读视频的第一帧。帧= readFrame (fileReader);%在第一帧提取所有车道点。lanePoints = gtdata.LaneBoundaries {1};在第一帧提取车辆边界框。vehicleBBox = gtdata.Vehicles {1};%叠加右车道点和车辆包围框。frame = insertMarker(frame, lanePoints{2}，“X”）;= insertObjectAnnotation帧(帧,“矩形”vehicleBBox,“汽车”）;%在第一帧显示地面真实数据。图imshow(框架)

%设置monoSensorHelper处理视频。monoCameraSensor = loaded.sensor;monoSensorHelper = helperMonoSensor (monoCameraSensor);用相同的时间矢量为测量创建新的时间表。测量=时间表(gtdata.Time);为保持车道边界和车辆数据设置时间表列。numFrames =地板(fileReader.FrameRate * fileReader.Duration);测量。LaneBoundaries = cell(numFrames, 2);测量。车辆检测= cell(numFrames, 1);gtdata。LanesInVehicleCoord = cell(numFrames, 2);%将视频倒带到t = 0，并创建一个帧索引来保持电流%框架。fileReader。CurrentTime = 0;frameIndex = 0;%循环播放视频文件，直到没有新的帧。而hasFrame(fileReader) frameIndex = frameIndex+1;帧= readFrame (fileReader);%使用processFrame方法计算检测。%这个方法可以被一个自定义的车道检测方法代替。detections = processFrame(monoSensorHelper，帧);%存储估计的车道边界和车辆检测。测量。(detections.leftEgoBoundary LaneBoundaries {frameIndex} =．..detections.rightEgoBoundary];测量。VehicleDetections {frameIndex} = detections.vehicleBoxes;%为了便于比较，转换地面真线指向%车辆坐标系。gtPointsThisFrame = gtdata.LaneBoundaries {frameIndex};车辆点数= cell(1, numel(gtpointthisframe));为ii = 1:numel(gtPointsThisFrame) vehiclePoints{ii} = imageToVehicle(monocamerassensor, gtPointsThisFrame{ii});结束%存储以车辆坐标表示的地面真值点。gtdata。LanesInVehicleCoord {frameIndex} = vehiclePoints;结束

%倒带视频到t = 0。fileReader。CurrentTime = 0;阅读视频的第一帧。帧= readFrame (fileReader);birdsEyeImage = transformImage (monoSensorHelper。BirdsEyeConfig,框架);%在鸟瞰图中提取第一帧右车道点。firstFrameVehiclePoints = gtdata.LanesInVehicleCoord {1};pointsInBEV = vehicleToImage (monoSensorHelper。BirdsEyeConfig, firstFrameVehiclePoints {2});%在框架上叠加点。birdsEyeImage = insertMarker(birdsEyeImage, pointsInBEV，“X”，“大小”6);在鸟瞰视图中显示变换后的点。图imshow (birdsEyeImage)

阈值= 0.25;%在米[numMatches, numMisses, numFalsePositives, assignments] =．..evaluateLaneBoundaries(测量。LaneBoundaries,．..gtdata。LanesInVehicleCoord,．..阈值);disp (['匹配数:'num2str (numMatches)]);disp ([“失误次数:”num2str (numMisses)]);disp ([“误报次数:”num2str (numFalsePositives)]);

averageDistance = helperComputeLaneStatistics(测量。LaneBoundaries,．..gtdata。LanesInVehicleCoord,．..作业,@mean);%绘制估算值与地面真实值之间的平均距离。图阻止(gtdata。时间,averageDistance) title(“估算值与地面真实值之间的平均距离”网格)在ylabel (“米距离”)传说(的左边界，“右边界”）

%使用单相机计算估计的点位置。[estVehiclePoints, estImagePoints] = helpergetcordingpoints (monocamerassensor，．..测量。LaneBoundaries,．..gtdata。LanesInVehicleCoord,．..作业);在测量时间表中增加估计车道。测量。EstimatedLanes = estImagePoints;测量。LanesInVehicleCoord = estVehiclePoints;%创建一个新的时间表，包含可视化所需的所有变量。名称= {“LanePoints”；“DetectedLanePoints”};类型= labelType ({“行”；“行”}）;= table(名称，类型，“VariableNames”, {“名字”，“类型”}）;(gtdata visualizeInFrame =时间表。时间,．..gtdata。LaneBoundaries,．..测量。EstimatedLanes,．..“VariableNames”、名称);创建groundTruth对象。数据源= groundTruthDataSource (videoName);dataToVisualize = groundTruth(dataSource, labelDefs, visualizeInFrame);%将上一节的所有结果保存到distanceData中。垫在%的临时文件夹。dataToLoad = [tempdir“distanceData.mat”];保存(dataToLoad,“monoSensorHelper”，“videoName”，“测量”，“gtdata”，“averageDistance”）;

函数stat = helperComputeLaneStatistics(estModels, gtPoints, assignments, fcnHandle) numFrames = length(estModels);%使左和右估计NaN默认表示缺乏%的数据。stat = NaN*ones(numFrames, 2);为frameInd = 1: numFrames%默认情况下使左右估计NaN。stat(frameInd，:) = NaN*ones(1,1); / /设置帧长度为idx = 1:长度(estModels {frameInd})%忽略假阳性分配。如果作业{frameInd} (idx) = = 0继续；结束% estModelInFrame中的第k个边界匹配第k个边界%元素在gtPointsInFrame中被赋值索引。这种模式下= estModels {frameInd} (idx);thisGT = gtPoints {frameInd}{作业{frameInd} (idx)};thisGTModel = driving.internal.piecewiseLinearBoundary (thisGT);如果意味着(thisGTModel.Points (:, 2)) > 0%左车道xPoints = thisGTModel.Points (: 1);yDist = 0(大小(xPoints));为index = 1:numel(xPoints) gtYPoints = thisGTModel.computeBoundaryModel(xPoints(index));testYPoints = thisModel.computeBoundaryModel (xPoints(指数));yDist(指数)= abs (testYPoints-gtYPoints);结束stat(frameInd, 1) = fcnHandle(yDist); / /删除帧其他的%右车道xPoints = thisGTModel.Points (: 1);yDist = 0(大小(xPoints));为index = 1:numel(xPoints) gtYPoints = thisGTModel.computeBoundaryModel(xPoints(index));testYPoints = thisModel.computeBoundaryModel (xPoints(指数));yDist(指数)= abs (testYPoints-gtYPoints);结束stat(frameInd, 2) = fcnHandle(yDist); / /删除帧结束结束结束结束

函数[vehiclePoints, imagePoints] = helpergetcordingpoints (monocamerassensor, estModels, gtPoints, assignments) numFrames = length(estModels);imagePoints = cell(numFrames, 1);车辆点数= cell(numFrames, 1);为frameInd = 1: numFrames如果isempty(assignments{frameInd}) imagePointsInFrame = [];vehiclePointsInFrame = [];其他的estModelInFrame = estModels {frameInd};gtPointsInFrame = gtPoints {frameInd};imagePointsInFrame = cell(length(estModelInFrame)， 1);vehiclePointsInFrame = cell(length(estModelInFrame)， 1);为idx = 1:长度(estModelInFrame)%忽略假阳性分配。如果imagePointsInFrame{idx} = [NaN NaN];继续；结束% estModelInFrame中的第k个边界匹配第k个边界%元素在gtPointsInFrame中被赋值索引。这种模式下= estModelInFrame (idx);thisGT = gtPointsInFrame{作业{frameInd} (idx)};xPoints = thisGT(:， 1); / /点击这里yPoints = thisModel.computeBoundaryModel (xPoints);vehiclePointsInFrame{idx} = [xPoints, yPoints];imagePointsInFrame{idx} = vehicleToImage(monocamerassensor， [xPoints yPoints]);结束结束vehiclePoints {frameInd} = vehiclePointsInFrame;imagePoints {frameInd} = imagePointsInFrame;%使imagePoints[]而不是{}遵守groundTruth对象。如果isempty(imagePoints{frameInd}) imagePoints{frameInd} = [];结束如果isempty(vehiclePoints{frameInd}) vehiclePoints{frameInd} = [];结束结束结束