Plot binaryOccupancyMap in Matlab App Designer

Hello, I am trying to plot the path plan plot generated from using the A* hybrid path plan planner. Howerver, not sure how to plot it on the Matlab App Designer.

I included the original code I used for path planning where I can generate a plot.

I believe the bit I still need to implement in the App Designer is the show(binaryOccupancyMap(localmatrixmap)) bit.

This is what I got so far in the Matlab App Designer:

        % Button pushed function: PlotpathplanButton
        function PlotpathplanButtonPushed(app, event)
            startPose = [50 30 pi/2]; %pi/2 is forward, pi is left, -pi/2 down, 0 right
            goalPose = [51 71 pi/2];

            % Create a binary occupancy map
            map = vehicleCostmap(app.localmatrixmap);%vehicleCostmap/binaryOccupancyMap

            %map.CollisionChecker = inflationCollisionChecker("InflationRadius",7);

            %Specify the collision-checking configuration in the CollisionChecker property of the costmap.
            %NOTE: InflationRadius and VehicleDimensions properties have been removed
            %since R2020b

            vehicleDims = vehicleDimensions(4.5,1.7); % 1 m long, 1 m wide
            numCircles = 3;
            ccConfig = inflationCollisionChecker(vehicleDims,numCircles);
            map.CollisionChecker = ccConfig;

            %Create a state space
            stateSpace = stateSpaceSE2;

            % Update state space bounds to be the same as map limits - used in binarymap not vehiclecostMap
            %stateSpace.StateBounds = [map.XWorldLimits;map.YWorldLimits;[-pi pi]];

            % Construct a state validator object using the statespace and map object
            validator = validatorVehicleCostmap(stateSpace,Map=map);
            %validator = validatorOccupancyMap(stateSpace,Map=map);

            % Set the validation distance for the validator
            validator.ValidationDistance = 0.01;

            % PROPERTIES - Assign the state validator object to the plannerHybridAStar object
            %planner = plannerHybridAStar(validator,AnalyticExpansionInterval=5,InterpolationDistance=5,NumMotionPrimitives=5,DirectionSwitchingCost=10), MotionPrimitiveLength=3;
            planner = plannerHybridAStar(validator,InterpolationDistance=2,AnalyticExpansionInterval=1,NumMotionPrimitives=5);
            %planner = plannerHybridAStar(validator);

            % Compute a path for the given start and goal poses
            %pathObj = plan(planner,startPose,goalPose,SearchMode='exhaustive');
            pathObj = plan(planner,startPose,goalPose);

            % Extract the path poses from the path object
            hybridPath = pathObj.States;

            show(binaryOccupancyMap(app.localmatrixmap))
            hold on

            % Start state
            scatter(app.UIAxes_3,startPose(1,1),startPose(1,2),"g","filled")
           
            % Goal state
            scatter(app.UIAxes_3,goalPose(1,1),goalPose(1,2),"r","filled")

            % Path
            legend(app.UIAxes_3,"Start Pose","Goal Pose","Hybrid A* Generated Path")
            legend(app.UIAxes_3,Location="northwest")
            set(app.UIAxes_3,'YDir','normal') %use 'reverse' if required to flip the x-axis

            %Plot path map - UIAXES_3
            plot(app.UIAxes_3,hybridPath(:,1),hybridPath(:,2),"r-",LineWidth=2)
            rotate3d(app.UIAxes_2,'on')

            %app.UIAxes_3.XLim = [35 70];
            %app.UIAxes_3.YLim = [35 70];

            view(app.UIAxes_3, 2);

            disp('Path map plotted');

            %Angles in degrees
            app.hybridPathInDegrees = hybridPath;
            app.hybridPathInDegrees(:,3) = 180/pi * hybridPath(:,3);

        end

[Walter Roberson]

show() does look plausible -- make sure you pass in the 'Parent' option to specify which axes to draw on.

See also the size requirements in the C/C++ section at https://www.mathworks.com/help/nav/ref/binaryoccupancymap.html

[JIAN-HONG YE ZHU]

I managed to plot it by using surf() and using scatter3 upon the surf() plot.

I think show() and binaryOccupancyMap() are not supported by the Matlab App Designer, at least from what I have looked at the UIAxes Matlab webpage?

[Walter Roberson]

Unfortunately I do not have that toolbox to test with.

Answer 1

[Praveen Reddy]

Hi JIAN-HONG,

I understand that you are trying to create an app to plot the path plan generated from A* hybrid path and unable to use ‘show()’ method to display the binary occupancy grid in the app UI axis. You can pass app UI axis object to the show() method as suggested below.

boMap=binaryOccupancyMap(app.localmatrixmap);
show(boMap, 'Parent', app.UIAxes);

Here is the modified call back function and few properties:

properties (Access = private)
    localmatrixmap;
    hybridPath;
    hybridPathInDegrees;
end

function PlotPathPlanButtonPushed(app, event)
    localstructmap=load("localmap.mat");
    app.localmatrixmap=cell2mat(struct2cell(localstructmap));
    tic

    startPose = [50 30 pi/2]; %pi/2 is forward, pi is left, -pi/2 down, 0 right
    goalPose = [51 71 pi/2]; 
    
    % Create a binary occupancy map
    map = vehicleCostmap(app.localmatrixmap);%vehicleCostmap/binaryOccupancyMap
    
    %map.CollisionChecker = inflationCollisionChecker("InflationRadius",7);
    
    %Specify the collision-checking configuration in the CollisionChecker property of the costmap.
    %NOTE: InflationRadius and VehicleDimensions properties have been removed
    %since R2020b
    
    vehicleDims = vehicleDimensions(4.5,1.7); % 1 m long, 1 m wide
    numCircles = 3;
    ccConfig = inflationCollisionChecker(vehicleDims,numCircles);
    map.CollisionChecker = ccConfig;
    
    %Create a state space
    stateSpace = stateSpaceSE2;
    
    % Update state space bounds to be the same as map limits - used in binarymap not vehiclecostMap
    %stateSpace.StateBounds = [map.XWorldLimits;map.YWorldLimits;[-pi pi]];
    
    % Construct a state validator object using the statespace and map object
    validator = validatorVehicleCostmap(stateSpace,Map=map);
    %validator = validatorOccupancyMap(stateSpace,Map=map);
    
    % Set the validation distance for the validator
    validator.ValidationDistance = 0.01;
    
    % PROPERTIES - Assign the state validator object to the plannerHybridAStar object
    %planner = plannerHybridAStar(validator,AnalyticExpansionInterval=5,InterpolationDistance=5
    % ,NumMotionPrimitives=5,DirectionSwitchingCost=10), MotionPrimitiveLength=3;
    planner = plannerHybridAStar(validator,InterpolationDistance=2,AnalyticExpansionInterval=1, ...
        NumMotionPrimitives=5);
    %planner = plannerHybridAStar(validator);
    
    % Compute a path for the given start and goal poses
    %pathObj = plan(planner,startPose,goalPose,SearchMode='exhaustive');
    pathObj = plan(planner,startPose,goalPose);
    
    % Extract the path poses from the path object
    app.hybridPath = pathObj.States;
    
    boMap=binaryOccupancyMap(app.localmatrixmap);
    show(boMap, 'Parent', app.UIAxes);

    hold(app.UIAxes,'on'); 
    % Start state
    scatter(app.UIAxes,startPose(1,1),startPose(1,2),"g","filled");
    hold(app.UIAxes,'on'); 
    % Goal state
    scatter(app.UIAxes,goalPose(1,1),goalPose(1,2),"r","filled");
    hold(app.UIAxes,'on'); 
    % Path
    plot(app.UIAxes,app.hybridPath(:,1),app.hybridPath(:,2),"r-",LineWidth=2);
    title(app.UIAxes,'Path Planning Map')
    legend(app.UIAxes,"Start Pose","Goal Pose","Hybrid A* Generated Path")
    legend(app.UIAxes,Location="northwest")
    set(app.UIAxes,'YDir','normal') %use 'reverse' if required to flip the x-axis
    
    %Angles in degrees
    app.hybridPathInDegrees = app.hybridPath;
    app.hybridPathInDegrees(:,3) = 180/pi * app.hybridPath(:,3);
    
    toc
end

Refer the attached “.mlapp” file which plots the path plan.

Please refer to the following MATLAB documentation to know more about "show()" method : https://www.mathworks.com/help/nav/ref/binaryoccupancymap.show.html