Ball finding vision system

Cuprins proiect Cum descarc?

Obiectivele proiectului .. 3
Tipuri de teste . 3
Unit Testing ... 4
Integration Testing . 5
System Testing 7
Concluzii .. 8

Extras din proiect Cum descarc?

Obiectivele proiectului
Proiectul este conceput pentru a ajuta dezvoltatorii de aplicatii pentru robotul roboRIO. Aplicatia simuleaza datele procesate de robot astfel se pot dezvolta si testa aplicatii de procesare a imaginilor WPILib fara a fi nevoie de un robot.
Pentru a demonstra functionalitatea librariei am creat un proiect de test care detecteaza bilele rosii si albastre dintr-o imagine.
.
Tipuri de teste:
Pentru acest proiect am ales sa fac urmatoarele tipuri de teste:
-Unit testing
-Integration testing
-System testing
Clasa testata:
public class ImageProcessor {
private IBallPipeline pipeline;
private INetworkTableWriter networkTableWriter;
private ExecutorService executor = Executors.newSingleThreadExecutor();
private Mat outputImage = new Mat();
private Future<?> processAsyncFuture;
/**
- ImageProcessor requires a pipeline to process and a network table writer to write
- results to.
- @param pipeline The pipeline to process
- @param networkTableWriter A network table writer to send results to
- /
public ImageProcessor(IBallPipeline pipeline, INetworkTableWriter networkTableWriter) {
if (pipeline == null) {
throw new IllegalArgumentException();
}
this.pipeline = pipeline;
this.networkTableWriter = networkTableWriter;
this.processAsyncFuture = null;
}
/**
- Process an image asynchronously. Call awaitProcessCompletion to wait for completion.
- You can only process one image at a time.
4
- @param inputImage The image to process
- /
public void processAsync(Mat inputImage)
{
if (processAsyncFuture != null) {
throw new IllegalAccessError("Only one process can be awaited at a time.");
}
// Hold on the the future...use the awaiter to wait for completion
processAsyncFuture = executor.submit(() -> {
// Apply the pipeline to the image.
pipeline.process(inputImage);
// Update network table
networkTableWriter.write();
return null;
});
}
- metoda pe care am testat-o este functia processAsync care face 2 lucruri: apeleaza pipelineul de lucru si scrie in tabela de network.
Unit Testing:
- am folosit frameworkul Mockito pentru a mockui interfetele de pipeline si networktabelwriter
- cu acest tip de test doresc sa verific ca se apeleaza metodele de process si write din functia processAsync
import org.junit.*;
import org.opencv.core.*;
import static org.mockito.Mockito.*;
public class ImageProcessorUnitTest
{
// This must be done in order to call opencv classes
static {
System.loadLibrary("opencv_java310");
}
@Test
public void itShouldProcessMyPipeline() {
IBallPipeline pipelineMock = mock(IBallPipeline.class);
INetworkTableWriter networkTableWriterMock = mock(INetworkTableWriter.class);
ImageProcessor imageProcessor = new ImageProcessor(pipelineMock, networkTableWriterMock);
5
Mat emptyImage = new Mat();
// Act
imageProcessor.processAsync(emptyImage);
imageProcessor.awaitProcessCompletion();
//Assert
verify(pipelineMock, times(1)).process(emptyImage);
}

Fisiere în arhivă (1):

 Ball finding vision system.pdf

Imagini din acest proiect Cum descarc?