Reference database and method for determining spectra using measurements from an LED color sensor, and method of generating a reference database
View Patent ↗To determine spectra, integrated multiple illuminant measurements from a non-fully illuminant populated color sensor may be converted into a fully populated spectral curve using a reference database. The reference database is partitioned into a plurality of clusters, and an appropriate centroid is determined for each cluster by, for example, vector quantization. Training samples that form the reference database may be assigned to the clusters by comparing the Euclidean distance between the centroids and the sample under consideration, and assigning each sample to the cluster having the centroid with the shortest Euclidean distance. When all training samples have been assigned, the resulting structure is stored as the reference database. When reconstructing the spectra for new measurements from the sensor, the Euclidean distances between actual color samples under measurement and each cluster centroid are measured. The spectra are then reconstructed using only the training samples from the cluster corresponding to the shortest Euclidean distance, resulting in improved speed and accuracy.
1. A method of partitioning a reference database for determining a reflectance spectrum, comprising;
establishing a plurality of clusters, each cluster initially being an empty set;
identifying, for each training sample of a plurality of training samples, a most appropriate cluster among the plurality of clusters: and
assigning each training sample to the most appropriate cluster, thereby filling the empty sets, each training sample correlating a reference spectrum with a corresponding plurality of normalized illuminant sensor outputs for reference colors.
2. The method according to claim 1 , wherein:
the establishing the plurality of clusters comprises establishing a plurality of cluster centroids; and
the identifying of the most appropriate cluster comprises obtaining, for each training sample, a Euclidean distance from the training sample to each of the cluster centroids, such that a plurality of Euclidean distances are determined for each training sample,
wherein the most appropriate cluster is determined to be the cluster associated with the cluster centroid having the shortest Euclidean distance among the plurality of Euclidean distances.
3. The method of claim 2 , further comprising:
obtaining an average distortion based on the shortest Euclidean distance for each training sample;
updating the cluster centroids to decrease the average distortion; and
re-identifying the most appropriate cluster for each training sample and re-assigning the training samples based on the updated cluster centroids.
4. The method according to claim 1 , wherein:
the establishing the plurality of clusters comprises establishing a plurality of cluster centroids, the cluster centroids being established through vector quantization.
5. A reference database partitioned by the method of claim 1 , the reference database being machine-readable.
6. A storage medium on which is recorded a program for implementing the method of claim 1 .
7. A method of partitioning a reference database for determining a reflectance spectrum, comprising:
establishing a plurality of clusters;
identifying, for each training sample of a plurality of training samples, a most appropriate cluster among the plurality of clusters; and
assigning each training sample to the most appropriate cluster, each training sample correlating a reference spectrum with a corresponding plurality of normalized illuminant sensor outputs for reference colors, wherein:
the establishing the plurality of clusters comprises establishing a plurality of cluster centroids; and
the identifying of the most appropriate cluster comprises obtaining, for each training sample, a Euclidean distance from the training sample to each of the cluster centroids, such that a plurality of Euclidean distances are determined for each training sample,
wherein the most appropriate cluster is determined to be the cluster associated with the cluster centroid having the shortest Euclidean distance among the plurality of Euclidean distances.
8. The method of claim 7 , further comprising:
obtaining an average distortion based on the shortest Euclidean distance for each training sample;
updating the cluster centroids to decrease the average distortion; and
re-identifying the most appropriate cluster for each training sample and re-assigning the training samples based on the updated cluster centroids.
9. A reference database partitioned by the method of claim 7 , the reference database being machine-readable.
10. A storage medium on which is recorded a program for implementing the method of claim 7 .