Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/usnistgov/NFIQ2/llms.txt

Use this file to discover all available pages before exploring further.

The NFIQ2::QualityMeasures::Algorithm class controls the computation of one or more quality measures. Functions in this namespace help organize and analyze computed algorithm objects.

​
Algorithm Class

namespace NFIQ2 { namespace QualityMeasures {
    class Algorithm;
}}
The Algorithm class is an abstract base class that controls the computation of one or more quality measures. Each specific quality measure algorithm (e.g., FrequencyDomainAnalysis, MinutiaeCount) is implemented as a subclass.
The Algorithm class is typically used through shared pointers (std::shared_ptr<NFIQ2::QualityMeasures::Algorithm>) returned by computation functions.

​
Key Characteristics

  • Encapsulation: Each algorithm object contains computed quality measure values, timing information, and metadata
  • Identifier-based: Each algorithm has a unique identifier from NFIQ2::Identifiers::QualityMeasureAlgorithms
  • Multiple measures: A single algorithm may compute multiple related quality measures (e.g., mean and standard deviation)
  • Performance tracking: Algorithms track computation time in milliseconds

​
Functions

​
getNativeQualityMeasureAlgorithms

std::unordered_map<std::string,
    std::shared_ptr<NFIQ2::QualityMeasures::Algorithm>>
getNativeQualityMeasureAlgorithms(
    const std::vector<std::shared_ptr<NFIQ2::QualityMeasures::Algorithm>>
        &algorithms);
Organize quality measure algorithms as a map with algorithm identifiers as keys.
​
algorithms
const std::vector<std::shared_ptr<NFIQ2::QualityMeasures::Algorithm>> &
required
Computed native quality measure algorithms, typically from computeNativeQualityMeasureAlgorithms().
​
return
std::unordered_map<std::string, std::shared_ptr<NFIQ2::QualityMeasures::Algorithm>>
Algorithms from the input vector organized in a map. Keys are algorithm identifiers from NFIQ2::Identifiers::QualityMeasureAlgorithms.
See also: Quality Measure Identifiers Example: 
#include <nfiq2_qualitymeasures.hpp>

auto algorithms = NFIQ2::QualityMeasures::computeNativeQualityMeasureAlgorithms(image);
auto algorithmMap = NFIQ2::QualityMeasures::getNativeQualityMeasureAlgorithms(algorithms);

// Access specific algorithm by identifier
auto fdaAlgorithm = algorithmMap["FrequencyDomainAnalysis"];

​
getNativeQualityMeasureAlgorithmSpeeds

std::unordered_map<std::string, double>
getNativeQualityMeasureAlgorithmSpeeds(
    const std::vector<std::shared_ptr<NFIQ2::QualityMeasures::Algorithm>>
        &algorithms);
Obtain the time elapsed during each quality measure algorithm’s computation.
​
algorithms
const std::vector<std::shared_ptr<NFIQ2::QualityMeasures::Algorithm>> &
required
Computed native quality measure algorithms, typically from computeNativeQualityMeasureAlgorithms().
​
return
std::unordered_map<std::string, double>
A map of quality measure algorithm identifiers to elapsed time in milliseconds. Keys are identifiers from NFIQ2::Identifiers::QualityMeasureAlgorithms.
Example: 
auto algorithms = NFIQ2::QualityMeasures::computeNativeQualityMeasureAlgorithms(image);
auto speeds = NFIQ2::QualityMeasures::getNativeQualityMeasureAlgorithmSpeeds(algorithms);

for (const auto &[algorithmId, milliseconds] : speeds) {
    std::cout << algorithmId << " computed in " 
              << milliseconds << " ms" << std::endl;
}
Performance Monitoring: 
// Profile algorithm performance across multiple images
std::unordered_map<std::string, std::vector<double>> performanceData;

for (const auto &image : images) {
    auto algorithms = NFIQ2::QualityMeasures::computeNativeQualityMeasureAlgorithms(image);
    auto speeds = NFIQ2::QualityMeasures::getNativeQualityMeasureAlgorithmSpeeds(algorithms);
    
    for (const auto &[alg, ms] : speeds) {
        performanceData[alg].push_back(ms);
    }
}

// Calculate average performance
for (const auto &[alg, times] : performanceData) {
    double avg = std::accumulate(times.begin(), times.end(), 0.0) / times.size();
    std::cout << alg << " average: " << avg << " ms" << std::endl;
}

​
Quality Measure Algorithm Types

NFIQ 2 includes the following native quality measure algorithms:
Analyzes the frequency of the sinusoid following the ridge-valley structure.Computes:
  • Mean of local frequency domain analysis
  • Standard deviation of local frequency domain analysis
  • Histogram of local quality values (10 bins)
Counts minutiae detected using the integrated FingerJet FX OSE feature extractor.Computes:
  • Total minutiae count
  • Minutiae count in center of mass (200x200 pixel region)
Assesses quality of detected minutiae based on local image characteristics.Computes:
  • Percentage of minutiae with quality based on image mean (0-0.5 range)
  • Percentage of minutiae with orientation certainty above 80
Measures the clarity of ridges and valleys in local regions.Computes:
  • Mean of local clarity score
  • Standard deviation of local clarity score
  • Histogram of local quality values (10 bins)
Analyzes gray level contrast across the image.Computes:
  • Average contrast (MU - arithmetic mean)
  • Average of average contrasts (MMB - mean of block means)
Measures the strength of energy concentration along dominant ridge flow orientation.Computes:
  • Mean of local orientation certainty level
  • Standard deviation of local orientation certainty level
  • Histogram of local quality values (10 bins)
Assesses ridge flow continuity based on orientation differences between neighboring blocks.Computes:
  • Mean of local orientation flow
  • Standard deviation of orientation flow
  • Histogram of local quality values (10 bins)
Analyzes the foreground (region of interest) area of the image.Computes:
  • Mean grayscale value of local regions in ROI
Measures coherence across the foreground area.Computes:
  • Sum of coherence values over ROI
  • Mean coherence value over ROI
Measures the consistency of ridge and valley widths.Computes:
  • Mean of local ridge valley uniformity
  • Standard deviation of local ridge valley uniformity
  • Histogram of local quality values (10 bins)

​
Usage Patterns

​
Basic Algorithm Access

NFIQ2::FingerprintImageData image(data, size, width, height, 0, 500);

// Compute all algorithms
auto algorithms = NFIQ2::QualityMeasures::computeNativeQualityMeasureAlgorithms(image);

// Organize by identifier for easy access
auto algorithmMap = NFIQ2::QualityMeasures::getNativeQualityMeasureAlgorithms(algorithms);

// Access specific algorithm
if (algorithmMap.find("MinutiaeCount") != algorithmMap.end()) {
    auto minutiaeAlg = algorithmMap["MinutiaeCount"];
    // Use algorithm...
}

​
Performance Profiling

// Compute algorithms with timing
auto algorithms = NFIQ2::QualityMeasures::computeNativeQualityMeasureAlgorithms(image);
auto speeds = NFIQ2::QualityMeasures::getNativeQualityMeasureAlgorithmSpeeds(algorithms);

// Find slowest algorithm
auto slowest = std::max_element(
    speeds.begin(), speeds.end(),
    [](const auto &a, const auto &b) { return a.second < b.second; }
);

std::cout << "Slowest: " << slowest->first 
          << " (" << slowest->second << " ms)" << std::endl;

​
Selective Processing

// Get list of all algorithm identifiers
auto algorithmIds = NFIQ2::QualityMeasures::getNativeQualityMeasureAlgorithmIDs();

std::cout << "Available algorithms:" << std::endl;
for (const auto &id : algorithmIds) {
    std::cout << "  - " << id << std::endl;
}

Quality Measures

Core functions for computing quality measures

Quality Measure Identifiers

View all available algorithm and measure identifiers

Build docs developers (and LLMs) love

Get started for freeTalk to us