Accelerating classification time in Hyperspectral Images

K-nearest neighbour (K-NN) is a supervised classification technique that is widely used in many fields of study to classify unknown queries based on some known information about the dataset. K-NN is known to be robust and simple to implement when dealing with data of small size. However its performance is slow when data is large and has high dimensions. Hyperspectral images, often collected from high altitudes, cover very large areas and consist of a large number of pixels, each having hundreds of spectral dimensions. We focus on one of the most popular algorithms for performing approximate search for large datasets based on the concept of locality-sensitive hashing (LSH) for Hyperspectral Image Processing, that allows us to quickly find similar entries in large databases. Our experiments show that LSH accelerates the classification time significantly without effecting the classification rates.