Fix Fourier transform dimension #131

Ajout d'une vérification sur les dimensions des images d'entrée.
Si ce n'est pas le cas, on ne réalise la FFT inverse
+ on génère une boite de dialogue

app/Operations/IFFTOp.cpp

@@ -31,6 +31,7 @@
 #include <cmath>
 #include <QGroupBox>
 #include <QCheckBox>
+#include <QMessageBox>
 
 using namespace std;
 using namespace imagein;
@@ -100,6 +101,11 @@ void IFFTOp::operator()(const imagein::Image_t<double>*, const map<const imagein
         const Image_t<double>* magnitudeImg = magtdImgBox->currentImage();
         const Image_t<double>* phaseImg = phaseImgBox->currentImage();
         if(magnitudeImg == NULL || phaseImg == NULL) return;
+        if(magnitudeImg->getHeight() != phaseImg->getHeight() || magnitudeImg->getWidth() != phaseImg->getWidth()){
+            QMessageBox::critical(nullptr, qApp->translate("IFFTOp", "Error : inverse FFT"),
+                                     qApp->translate("IFFTOp", "Magnitude and phase images have different size"));
+            return;
+        }
         unsigned int width = min(nearestUpPower2(magnitudeImg->getWidth()), nearestUpPower2(phaseImg->getWidth()));
         unsigned int height = min(nearestUpPower2(magnitudeImg->getHeight()), nearestUpPower2(phaseImg->getHeight()));
         unsigned int channels = min(magnitudeImg->getNbChannels(), phaseImg->getNbChannels());
@@ -152,6 +158,11 @@ void IFFTOp::operator()(const imagein::Image_t<double>*, const map<const imagein
         const Image_t<double>* realImg = realImgBox->currentImage();
         const Image_t<double>* imagImg = imagImgBox->currentImage();
         if(realImg == NULL || imagImg == NULL) return;
+        if(realImg->getHeight() != imagImg->getHeight() || realImg->getWidth() != imagImg->getWidth()){
+            QMessageBox::critical(nullptr, qApp->translate("IFFTOp", "Error : inverse FFT"),
+                                     qApp->translate("IFFTOp", "Real and imaginary images have different size"));
+            return;
+        }
         unsigned int width = min(nearestUpPower2(realImg->getWidth()), nearestUpPower2(imagImg->getWidth()));
         unsigned int height = min(nearestUpPower2(realImg->getHeight()), nearestUpPower2(imagImg->getHeight()));
         unsigned int channels = min(realImg->getNbChannels(), imagImg->getNbChannels());