diff --git a/Modules/Core/Common/include/itkVector.h b/Modules/Core/Common/include/itkVector.h
index 02fd8fec328..8a5e1cdf8af 100644
--- a/Modules/Core/Common/include/itkVector.h
+++ b/Modules/Core/Common/include/itkVector.h
@@ -90,7 +90,8 @@ class ITK_TEMPLATE_EXPORT Vector : public FixedArray<T, VVectorDimension>
     return VVectorDimension;
   }
 
-  /** Set a vnl_vector_ref referencing the same memory block. */
+  /** Copy values from the vnl_vector input to the internal memory block.  The minimum of
+   *  VVectorDimension and vnl_vector::size() elements are copied. */
   void
   SetVnlVector(const vnl_vector<T> &);
 
diff --git a/Modules/Core/Common/include/itkVector.hxx b/Modules/Core/Common/include/itkVector.hxx
index 408a9d8cb4f..8d2fca35141 100644
--- a/Modules/Core/Common/include/itkVector.hxx
+++ b/Modules/Core/Common/include/itkVector.hxx
@@ -155,7 +155,8 @@ template <typename T, unsigned int TVectorDimension>
 void
 Vector<T, TVectorDimension>::SetVnlVector(const vnl_vector<T> & v)
 {
-  for (unsigned int i = 0; i < v.size(); ++i)
+  const unsigned int elements_to_copy = std::min<unsigned int>(TVectorDimension, v.size());
+  for (unsigned int i = 0; i < elements_to_copy; ++i)
   {
     (*this)[i] = v(i);
   }
diff --git a/Modules/Segmentation/LevelSets/include/itkParallelSparseFieldLevelSetImageFilter.hxx b/Modules/Segmentation/LevelSets/include/itkParallelSparseFieldLevelSetImageFilter.hxx
index 255d2afac5c..cd8955db488 100644
--- a/Modules/Segmentation/LevelSets/include/itkParallelSparseFieldLevelSetImageFilter.hxx
+++ b/Modules/Segmentation/LevelSets/include/itkParallelSparseFieldLevelSetImageFilter.hxx
@@ -768,14 +768,14 @@ void
 ParallelSparseFieldLevelSetImageFilter<TInputImage, TOutputImage>::ThreadedAllocateData(ThreadIdType ThreadId)
 {
   static constexpr float SAFETY_FACTOR = 4.0;
-  unsigned int           i, j;
 
   m_Data[ThreadId].m_Semaphore[0] = 0;
   m_Data[ThreadId].m_Semaphore[1] = 0;
 
+  const std::size_t bufferLayerSize = 2 * m_NumberOfLayers + 1;
   // Allocate the layers for the sparse field.
-  m_Data[ThreadId].m_Layers.reserve(2 * m_NumberOfLayers + 1);
-  for (i = 0; i < 2 * static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
+  m_Data[ThreadId].m_Layers.reserve(bufferLayerSize);
+  for (unsigned int i = 0; i < 2 * static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
   {
     m_Data[ThreadId].m_Layers.push_back(LayerType::New());
   }
@@ -786,12 +786,13 @@ ParallelSparseFieldLevelSetImageFilter<TInputImage, TOutputImage>::ThreadedAlloc
   }
 
   // Layers used as buffers for transferring pixels during load balancing
-  m_Data[ThreadId].m_LoadTransferBufferLayers = new LayerListType[2 * m_NumberOfLayers + 1];
-  for (i = 0; i < 2 * static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
+
+  m_Data[ThreadId].m_LoadTransferBufferLayers = new LayerListType[bufferLayerSize];
+  for (unsigned int i = 0; i < 2 * static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
   {
     m_Data[ThreadId].m_LoadTransferBufferLayers[i].reserve(m_NumOfWorkUnits);
 
-    for (j = 0; j < m_NumOfWorkUnits; ++j)
+    for (unsigned int j = 0; j < m_NumOfWorkUnits; ++j)
     {
       m_Data[ThreadId].m_LoadTransferBufferLayers[i].push_back(LayerType::New());
     }
@@ -803,14 +804,13 @@ ParallelSparseFieldLevelSetImageFilter<TInputImage, TOutputImage>::ThreadedAlloc
 
   // The SAFETY_FACTOR simple ensures that the number of nodes created
   // is larger than those required to start with for each thread.
-  auto nodeNum =
-    static_cast<unsigned int>(SAFETY_FACTOR * m_Layers[0]->Size() * (2 * m_NumberOfLayers + 1) / m_NumOfWorkUnits);
+  auto nodeNum = static_cast<unsigned int>(SAFETY_FACTOR * m_Layers[0]->Size() * (bufferLayerSize) / m_NumOfWorkUnits);
 
   m_Data[ThreadId].m_LayerNodeStore->Reserve(nodeNum);
   m_Data[ThreadId].m_RMSChange = m_ValueZero;
 
   // UpLists and Downlists
-  for (i = 0; i < 2; ++i)
+  for (unsigned int i = 0; i < 2; ++i)
   {
     m_Data[ThreadId].UpList[i] = LayerType::New();
     m_Data[ThreadId].DownList[i] = LayerType::New();
@@ -824,27 +824,23 @@ ParallelSparseFieldLevelSetImageFilter<TInputImage, TOutputImage>::ThreadedAlloc
   // for the Downlists
   m_Data[ThreadId].m_InterNeighborNodeTransferBufferLayers[1] = new LayerPointerType *[m_NumberOfLayers + 1];
 
-  for (i = 0; i < static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
+  for (unsigned int i = 0; i < static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
   {
     m_Data[ThreadId].m_InterNeighborNodeTransferBufferLayers[0][i] = new LayerPointerType[m_NumOfWorkUnits];
     m_Data[ThreadId].m_InterNeighborNodeTransferBufferLayers[1][i] = new LayerPointerType[m_NumOfWorkUnits];
   }
 
-  for (i = 0; i < static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
+  for (unsigned int i = 0; i < static_cast<unsigned int>(m_NumberOfLayers) + 1; ++i)
   {
-    for (j = 0; j < m_NumOfWorkUnits; ++j)
+    for (unsigned int j = 0; j < m_NumOfWorkUnits; ++j)
     {
       m_Data[ThreadId].m_InterNeighborNodeTransferBufferLayers[0][i][j] = LayerType::New();
       m_Data[ThreadId].m_InterNeighborNodeTransferBufferLayers[1][i][j] = LayerType::New();
     }
   }
 
-  // Local histogram for every thread (used during Iterate() )
-  m_Data[ThreadId].m_ZHistogram = new int[m_ZSize];
-  for (i = 0; i < m_ZSize; ++i)
-  {
-    m_Data[ThreadId].m_ZHistogram[i] = 0;
-  }
+  // Local histogram for every thread (used during Iterate()), initialized to zeros.
+  m_Data[ThreadId].m_ZHistogram = new int[m_ZSize]();
 
   // Every thread must have its own copy of the GlobalData struct.
   m_Data[ThreadId].globalData = this->GetDifferenceFunction()->GetGlobalDataPointer();