public void BKTree_should_QueryBestMatchesBelowGivenThreshold()
{
BKTree <TestNode> tree = new BKTree <TestNode>();
TestNode search = new TestNode(new int[] { 399, 400, 400 });
TestNode best1 = new TestNode(41, new int[] { 400, 400, 400 });
TestNode best2 = new TestNode(42, new int[] { 403, 403, 403 });
TestNode best3 = new TestNode(43, new int[] { 406, 406, 406 });
tree.add(new TestNode(1, new int[] { 100, 100, 100 }));
tree.add(new TestNode(2, new int[] { 200, 200, 200 }));
tree.add(new TestNode(3, new int[] { 300, 300, 300 }));
tree.add(best1);
tree.add(best2);
tree.add(new TestNode(5, new int[] { 500, 500, 500 }));
Dictionary <TestNode, Int32> results;
// Query for match within distance of 1 (best1 is only expected result)
results = tree.query(search, 1);
Assert.Equal(1, results.Count);
Assert.Equal(1, DistanceMetric.calculateLeeDistance(search.Data, best1.Data));
Assert.Equal(1, results.Values.ElementAt(0));
Assert.Equal(41, results.Keys.ElementAt(0).Id);
Assert.Equal(best1.Data, results.Keys.ElementAt(0).Data);
// Query for match within distance of 10 (best1 & best2 are expected results)
tree.add(best3); // exercise adding another node after already queried
results = tree.query(search, 10);
Assert.Equal(2, results.Count);
Assert.Equal(1, DistanceMetric.calculateLeeDistance(search.Data, best1.Data));
Assert.Equal(10, DistanceMetric.calculateLeeDistance(search.Data, best2.Data));
Assert.True(results.Contains(new KeyValuePair <TestNode, int>(best1, 1)));
Assert.True(results.Contains(new KeyValuePair <TestNode, int>(best2, 10)));
// Query for matches within distance of 20 (best1, best2 & best3 are expected results)
results = tree.query(search, 20);
Assert.Equal(3, results.Count);
Assert.Equal(1, DistanceMetric.calculateLeeDistance(search.Data, best1.Data));
Assert.Equal(10, DistanceMetric.calculateLeeDistance(search.Data, best2.Data));
Assert.Equal(19, DistanceMetric.calculateLeeDistance(search.Data, best3.Data));
Assert.True(results.Contains(new KeyValuePair <TestNode, int>(best1, 1)));
Assert.True(results.Contains(new KeyValuePair <TestNode, int>(best2, 10)));
Assert.True(results.Contains(new KeyValuePair <TestNode, int>(best3, 19)));
}
/*
* To use BKTree:
* 1. Create a class dervied from BKTreeNode
* 2. Add a member variable of your data to be sorted / retrieved
* 3. Override the calculateDistance method to calculate the distance metric
* between two nodes for the data to be sorted / retrieved.
* 4. Instantiate a BKTree with the type name of the class created in (1).
*/
static void Main(string[] args)
{
/*
* NOTE: More comprehensive examples of BK-Tree methods in unit tests
*/
// Exercise static distance metric methods -- just because
Console.WriteLine(
DistanceMetric.calculateHammingDistance(
new byte[] { 0xEF, 0x35, 0x20 },
new byte[] { 0xAD, 0x13, 0x87 }));
Console.WriteLine(
DistanceMetric.calculateLeeDistance(
new int[] { 196, 105, 48 },
new int[] { 201, 12, 51 }));
Console.WriteLine(
DistanceMetric.calculateLevenshteinDistance(
"kitten",
"sitting"));
// Create BKTree with derived node class from top of file
BKTree <ExampleNodeRecord> tree = new BKTree <ExampleNodeRecord>();
// Add some nodes
tree.add(new ExampleNodeRecord(1, new int[] { 100, 200, 300 }));
tree.add(new ExampleNodeRecord(2, new int[] { 110, 210, 310 }));
tree.add(new ExampleNodeRecord(3, new int[] { 120, 220, 320 }));
tree.add(new ExampleNodeRecord(4, new int[] { 130, 230, 330 }));
tree.add(new ExampleNodeRecord(5, new int[] { 140, 240, 340 }));
// Get best node from our tree with best distance
Dictionary <ExampleNodeRecord, Int32> results =
tree.findBestNodeWithDistance(
new ExampleNodeRecord(new int[] { 103, 215, 303 }));
// Get best nodes below threshold
results = tree.query(
new ExampleNodeRecord(new int[] { 103, 215, 303 }),
10); // arbitrary threshold
// Dictionaries don't print well; so invent your own handy print routine
}
public List <ImageRecord> QueryImage(string queryImagePath, object argument = null)
{
List <ImageRecord> rtnImageList = new List <ImageRecord>();
CEDD_Descriptor.CEDD cedd = new CEDD_Descriptor.CEDD();
int goodMatchDistance = 35;
if (argument != null && argument is Int32)
{
goodMatchDistance = (int)argument;
}
double[] queryCeddDiscriptor;
using (Bitmap bmp = new Bitmap(Image.FromFile(queryImagePath)))
{
queryCeddDiscriptor = cedd.Apply(bmp);
}
Stopwatch sw = Stopwatch.StartNew();
BKTree <CEDDTreeNode> ceddTree = null;
if (!CacheHelper.Get <BKTree <CEDDTreeNode> >("CeddIndexTree", out ceddTree))
{
CEDDRepository <BKTree <CEDDTreeNode> > repo = new CEDDRepository <BKTree <CEDDTreeNode> >();
ceddTree = repo.Load();
if (ceddTree == null)
{
throw new InvalidOperationException("Please index CEDD with BK-Tree before querying the Image");
}
CacheHelper.Add <BKTree <CEDDTreeNode> >(ceddTree, "CeddIndexTree");
}
sw.Stop();
Debug.WriteLine("Load tooked {0} ms", sw.ElapsedMilliseconds);
CEDDTreeNode queryNode = new CEDDTreeNode
{
Id = 0,
ImagePath = queryImagePath,
CEDDDiscriptor = queryCeddDiscriptor
};
sw.Reset(); sw.Start();
Dictionary <CEDDTreeNode, Int32> result = ceddTree.query(queryNode, goodMatchDistance);
sw.Stop();
Debug.WriteLine("Query tooked {0} ms", sw.ElapsedMilliseconds);
foreach (KeyValuePair <CEDDTreeNode, Int32> ceddNode in result)
{
ImageRecord rec = new ImageRecord
{
Id = ceddNode.Key.Id,
ImageName = ceddNode.Key.ImageName,
ImagePath = ceddNode.Key.ImagePath,
Distance = ceddNode.Value
};
rtnImageList.Add(rec);
}
rtnImageList = rtnImageList.OrderBy(x => x.Distance).ToList();
return(rtnImageList);
}
/*
* To use BKTree:
* 1. Create a class dervied from BKTreeNode
* 2. Add a member variable of your data to be sorted / retrieved
* 3. Override the calculateDistance method to calculate the distance metric
* between two nodes for the data to be sorted / retrieved.
* 4. Instantiate a BKTree with the type name of the class created in (1).
*/
static void Main(string[] args)
{
/*
* NOTE: More comprehensive examples of BK-Tree methods in unit tests
*/
// Exercise static distance metric methods -- just because
Console.WriteLine(
DistanceMetric.calculateHammingDistance(
new byte[] { 0xEF, 0x35, 0x20 },
new byte[] { 0xAD, 0x13, 0x87 }));
Console.WriteLine(
DistanceMetric.calculateLeeDistance(
new int[] { 196, 105, 48 },
new int[] { 201, 12, 51 }));
Console.WriteLine(
DistanceMetric.calculateLevenshteinDistance(
"kitten",
"sitting"));
// Create BKTree with derived node class from top of file
BKTree<ExampleNodeRecord> tree = new BKTree<ExampleNodeRecord>();
// Add some nodes
tree.add( new ExampleNodeRecord( 1, new int[] {100,200,300}) );
tree.add( new ExampleNodeRecord( 2, new int[] {110,210,310}) );
tree.add( new ExampleNodeRecord( 3, new int[] {120,220,320}) );
tree.add( new ExampleNodeRecord( 4, new int[] {130,230,330}) );
tree.add( new ExampleNodeRecord( 5, new int[] {140,240,340}) );
// Get best node from our tree with best distance
Dictionary<ExampleNodeRecord, Int32> results =
tree.findBestNodeWithDistance(
new ExampleNodeRecord( new int[] { 103, 215, 303 }) );
// Get best nodes below threshold
results = tree.query(
new ExampleNodeRecord(new int[] { 103, 215, 303 }),
10 ); // arbitrary threshold
// Dictionaries don't print well; so invent your own handy print routine
}
public void BKTree_should_QueryBestMatchesBelowGivenThreshold()
{
BKTree<TestNode> tree = new BKTree<TestNode>();
TestNode search = new TestNode(new int[] { 399, 400, 400 });
TestNode best1 = new TestNode(41, new int[] { 400, 400, 400 });
TestNode best2 = new TestNode(42, new int[] { 403, 403, 403 });
TestNode best3 = new TestNode(43, new int[] { 406, 406, 406 });
tree.add(new TestNode(1, new int[] { 100, 100, 100 }));
tree.add(new TestNode(2, new int[] { 200, 200, 200 }));
tree.add(new TestNode(3, new int[] { 300, 300, 300 }));
tree.add(best1);
tree.add(best2);
tree.add(new TestNode(5, new int[] { 500, 500, 500 }));
Dictionary<TestNode, Int32> results;
// Query for match within distance of 1 (best1 is only expected result)
results = tree.query(search, 1);
Assert.Equal(1, results.Count);
Assert.Equal(1, DistanceMetric.calculateLeeDistance(search.Data, best1.Data));
Assert.Equal(1, results.Values.ElementAt(0));
Assert.Equal(41, results.Keys.ElementAt(0).Id);
Assert.Equal(best1.Data, results.Keys.ElementAt(0).Data);
// Query for match within distance of 10 (best1 & best2 are expected results)
tree.add(best3); // exercise adding another node after already queried
results = tree.query(search, 10);
Assert.Equal(2, results.Count);
Assert.Equal(1, DistanceMetric.calculateLeeDistance(search.Data, best1.Data));
Assert.Equal(10, DistanceMetric.calculateLeeDistance(search.Data, best2.Data));
Assert.True(results.Contains(new KeyValuePair<TestNode, int>(best1, 1)));
Assert.True(results.Contains(new KeyValuePair<TestNode, int>(best2, 10)));
// Query for matches within distance of 20 (best1, best2 & best3 are expected results)
results = tree.query(search, 20);
Assert.Equal(3, results.Count);
Assert.Equal(1, DistanceMetric.calculateLeeDistance(search.Data, best1.Data));
Assert.Equal(10, DistanceMetric.calculateLeeDistance(search.Data, best2.Data));
Assert.Equal(19, DistanceMetric.calculateLeeDistance(search.Data, best3.Data));
Assert.True(results.Contains(new KeyValuePair<TestNode, int>(best1, 1)));
Assert.True(results.Contains(new KeyValuePair<TestNode, int>(best2, 10)));
Assert.True(results.Contains(new KeyValuePair<TestNode, int>(best3, 19)));
}