【图像配准】基于双目视觉sift特征点实现图像配准附Matlab代码

双目视觉SIFT特征在图像配准中的应用与MATLAB实现

原创于 2024-04-27 23:37:33 发布 · 1.2k 阅读

15 ·

CC 4.0 BY-SA版权

文章标签：

#matlab

图像处理专栏收录该内容

871 篇文章

订阅专栏

本文介绍了双目视觉中利用SIFT特征点进行图像配准的方法，包括特征提取、匹配、立体匹配和图像对齐过程。强调了这种方法的鲁棒性、精度和实时性能，并提供了MATLAB代码示例，适用于机器人导航、医学图像等领域。

✅作者简介：热爱科研的Matlab仿真开发者，修心和技术同步精进，代码获取、论文复现及科研仿真合作可私信。

🍎个人主页：Matlab科研工作室

🍊个人信条：格物致知。

更多Matlab完整代码及仿真定制内容点击👇

智能优化算法神经网络预测雷达通信无线传感器电力系统

信号处理图像处理路径规划元胞自动机无人机

物理应用机器学习

🔥 内容介绍

图像配准是指将两幅或多幅图像在空间上进行对齐，使它们在同一个坐标系下进行比较和分析。双目视觉是实现图像配准的一种常见方法，它利用两个摄像头采集同一场景的不同视角图像，通过特征点匹配和立体匹配算法来计算两幅图像之间的相对位姿，从而实现图像配准。

SIFT (Scale-Invariant Feature Transform) 是一种尺度不变的特征描述子，它对图像的旋转、缩放、亮度变化等具有较好的鲁棒性。在双目视觉图像配准中，SIFT特征点可以用来匹配两幅图像中的对应点，从而获得两幅图像之间的匹配关系。

基于双目视觉SIFT特征点实现图像配准的步骤如下：

特征提取：对两幅图像分别提取SIFT特征点。
特征匹配：利用特征描述子进行特征匹配，找到两幅图像中对应点的匹配关系。
立体匹配：根据匹配关系和双目相机的参数，计算两幅图像之间的相对位姿。
图像配准：根据计算出的相对位姿，将其中一幅图像进行空间变换，使其与另一幅图像对齐。

基于双目视觉SIFT特征点实现图像配准具有以下优点：

鲁棒性强： SIFT特征点对图像的旋转、缩放、亮度变化等具有较好的鲁棒性。
精度高：双目视觉可以提供精确的深度信息，提高图像配准的精度。
实时性好： SIFT特征点提取和匹配算法可以快速完成，满足实时应用的需求。

总而言之，基于双目视觉SIFT特征点实现图像配准是一种高效、鲁棒的图像配准方法，它在许多领域都得到了广泛的应用，例如机器人导航、三维重建、医学图像配准等。

📣 部分代码

%%Pre Process Datads=datastore("LakeDATA\*Quality*.csv","TextType","string","DatetimeType","text");Week1=read(ds);Week1.Date= datetime(Week1.Date,'InputFormat','yyyy-MM-dd');Week1.Date.Format='dd/MM/uuuu';Week1 = rmmissing(Week1,"DataVariables",width(Week1));%% Set up the Import Options and import the dataopts = spreadsheetImportOptions("NumVariables", 3);% Specify sheet and rangeopts.Sheet = "Sheet1";opts.DataRange = "A2:C6";% Specify column names and typesopts.VariableNames = ["Parameters", "StandardValue", "IdealValue"];opts.VariableTypes = ["string", "double", "double"];% Specify variable propertiesopts = setvaropts(opts, "Parameters", "WhitespaceRule", "preserve");opts = setvaropts(opts, "Parameters", "EmptyFieldRule", "auto");% Import the dataQualityIndex = readtable("C:\Users\celeste\Desktop\fer\Proyecto Fase I\Matlab\QualityIndex.xlsx", opts, "UseExcel", false)%% Clear temporary variablesclear opts%% Analyze DataWeekdata=Week1{:,3:7};meanWeek= mean(Weekdata,'omitnan');medianWeek= median(Weekdata,'omitnan');modeWeek=mode(Weekdata);varianceWeek=var(Weekdata,'omitnan');stdWeek = std(Weekdata, 'omitnan');maxWeek= max(Weekdata);minWeek= min(Weekdata);rangeWeek= range(Weekdata);departureWeek= Weekdata-meanWeek;normalizedWeek = (Weekdata - meanWeek) ./ stdWeek;analysisTable = table(meanWeek', medianWeek', modeWeek', varianceWeek', stdWeek', maxWeek', minWeek', rangeWeek');analysisTable.Properties.VariableNames = {'Mean', 'Median', 'Mode', 'Variance', 'Standard Deviation', 'Max', 'Min', 'Range'};analysisTable.Properties.RowNames = {'Temperature', 'pH', 'TDS', 'Turbulence', 'ORP'}WeekdataTable = array2table(Weekdata, 'VariableNames', {'Temperature', 'pH', 'TDS', 'Turbidity', 'ORP'});k=sum(1./QualityIndex.StandardValue);k=1/k;Wi=(k ./QualityIndex.StandardValue);Qs=(QualityIndex.StandardValue-QualityIndex.IdealValue);for i = 1:size(WeekdataTable, 2)    parameter = QualityIndex.Parameters{i};    standard_value = QualityIndex.StandardValue(i);    ideal_value = QualityIndex.IdealValue(i);    idx = strcmp(QualityIndex.Parameters, parameter);    StandardTable(:, i) = abs(standard_value-WeekdataTable(:, i));    IdealTable(:,i)=abs(WeekdataTable(:,i)-ideal_value);endQt = IdealTable{:,:};StandardMatrix = table2array(StandardTable);Qi = 100*(Qt ./ StandardMatrix);QiWi=Qi.*Wi';WQI=sum(QiWi,2);Week1 = addvars(Week1, WQI, 'After', 'ORP', 'NewVariableNames', 'WQI')yfit = Ensemblemdl.predictFcn(Week1) function [trainedModel, validationRMSE] = trainRegressionModel(trainingData)% [trainedModel, validationRMSE] = trainRegressionModel(trainingData)% Returns a trained regression model and its RMSE. This code recreates the% model trained in Regression Learner app. Use the generated code to% automate training the same model with new data, or to learn how to% programmatically train models.%%  Input:%      trainingData: A table containing the same predictor and response%       columns as those imported into the app.%%%  Output:%      trainedModel: A struct containing the trained regression model. The%       struct contains various fields with information about the trained%       model.%%      trainedModel.predictFcn: A function to make predictions on new data.%%      validationRMSE: A double representing the validation RMSE. In the%       app, the Models pane displays the validation RMSE for each model.%% Use the code to train the model with new data. To retrain your model,% call the function from the command line with your original data or new% data as the input argument trainingData.%% For example, to retrain a regression model trained with the original data% set T, enter:%   [trainedModel, validationRMSE] = trainRegressionModel(T)%% To make predictions with the returned 'trainedModel' on new data T2, use%   yfit = trainedModel.predictFcn(T2)%% T2 must be a table containing at least the same predictor columns as used% during training. For details, enter:%   trainedModel.HowToPredict% Auto-generated by MATLAB on 13-Apr-2024 02:26:21% Extract predictors and response% This code processes the data into the right shape for training the% model.inputTable = trainingData;predictorNames = {'Temperature', 'pH', 'TDS', 'ORP', 'WQI'};predictors = inputTable(:, predictorNames);response = inputTable.Turbulency;isCategoricalPredictor = [false, false, false, false, false];% Train a regression model% This code specifies all the model options and trains the model.template = templateTree(...    'MinLeafSize', 8, ...    'NumVariablesToSample', 'all');regressionEnsemble = fitrensemble(...    predictors, ...    response, ...    'Method', 'Bag', ...    'NumLearningCycles', 30, ...    'Learners', template);% Create the result struct with predict functionpredictorExtractionFcn = @(t) t(:, predictorNames);ensemblePredictFcn = @(x) predict(regressionEnsemble, x);trainedModel.predictFcn = @(x) ensemblePredictFcn(predictorExtractionFcn(x));% Add additional fields to the result structtrainedModel.RequiredVariables = {'Temperature', 'pH', 'TDS', 'ORP', 'WQI'};trainedModel.RegressionEnsemble = regressionEnsemble;trainedModel.About = 'This struct is a trained model exported from Regression Learner R2023b.';trainedModel.HowToPredict = sprintf('To make predictions on a new table, T, use: \n  yfit = c.predictFcn(T) \nreplacing ''c'' with the name of the variable that is this struct, e.g. ''trainedModel''. \n \nThe table, T, must contain the variables returned by: \n  c.RequiredVariables \nVariable formats (e.g. matrix/vector, datatype) must match the original training data. \nAdditional variables are ignored. \n \nFor more information, see <a href="matlab:helpview(fullfile(docroot, ''stats'', ''stats.map''), ''appregression_exportmodeltoworkspace'')">How to predict using an exported model</a>.');% Extract predictors and response% This code processes the data into the right shape for training the% model.inputTable = trainingData;predictorNames = {'Temperature', 'pH', 'TDS', 'ORP', 'WQI'};predictors = inputTable(:, predictorNames);response = inputTable.Turbulency;isCategoricalPredictor = [false, false, false, false, false];% Perform cross-validationpartitionedModel = crossval(trainedModel.RegressionEnsemble, 'KFold', 5);% Compute validation predictionsvalidationPredictions = kfoldPredict(partitionedModel);% Compute validation RMSEvalidationRMSE = sqrt(kfoldLoss(partitionedModel, 'LossFun', 'mse'));end