Particle measuring device and particle measuring method
Inventors
Saito, Mitsuaki • SUGAHARA, Reiko • YOSHIKAWA, Yuka
Assignees
Publication Number
US-12181398-B2
Publication Date
2024-12-31
Expiration Date
2041-12-21
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Abstract
A particle measuring device includes: a detection unit that detects scattered light generated due to interaction between a particle contained in a liquid sample and light incident thereon, and converts the detected scattered light into a signal; an addition unit that performs a predetermined number of parallel processing on the signal to add the predetermined number of uncorrelated noises thereto and outputs the resulting signals; a binarization unit that binarizes the resulting signals using a binarization threshold set in accordance with the liquid sample, and outputs the binarized signals; a calculation unit that calculates and outputs a value based on the binarized signals; a filter unit that passes a predetermined frequency component of the output of the calculation unit; and a determination unit that determines that the particle is present when an output of the filter unit exceeds a predetermined particle threshold.
Core Innovation
The invention provides a particle measuring device and method that improve detection accuracy and counting efficiency of fine particles in a liquid sample by using stochastic resonance signal processing tailored to the liquid's characteristics. The device detects scattered light generated by particles interacting with incident light and processes the signal through parallel addition of multiple uncorrelated noises, followed by binarization using a threshold set according to the liquid sample. This adaptive thresholding accounts for noise variations derived from the sample's scattered light, which is significantly larger and variable compared to air, enabling reliable detection of weak particle signals even below 30 nm.
The invention addresses the problem that conventional stochastic resonance methods employing a fixed binarization threshold fail to accurately detect particles in liquid samples due to the strong and variable background scattered light noise inherent to liquids. Manual adjustment of the binarization threshold for each sample is time-consuming and prone to instability. Furthermore, the preparation of multiple uncorrelated noises for parallel processing is resource-consuming. The invention's techniques for automatic threshold calculation based on characteristic noise measures like RMS or DC values of the scattered light component, and efficient noise generation by shifting start points in stored noise data, overcome these challenges.
By lowering the level of background light-derived signals and enhancing the relative prominence of particle-derived signals through these adaptive and efficient processing steps, the invention significantly improves measurement capability, including reducing the minimum measurable particle size down to about 20 nm or less. This allows the device to cope with variables such as refractive index, concentration, and temperature changes in liquid samples without manual intervention, thus leading to more stable, accurate, and efficient particle detection and counting.
Claims Coverage
The claims include five independent claims that cover a particle measuring device, a particle measuring method, and a signal processing program, each incorporating key features of adaptive stochastic resonance processing for particle detection in liquid samples.
Particle measuring device employing adaptive stochastic resonance processing
The device comprises an optical detector converting scattered light from particles in a liquid sample into a signal, and a processor implementing: parallel addition of multiple uncorrelated noises to the signal; binarization of resulting signals using a binarization threshold set according to the liquid sample; calculation of a value based on binarized signals; filtering to pass predetermined frequency components; determination of particle presence if the filtered output exceeds a preset particle threshold set according to the minimum measurable particle size; and calculation of the binarization threshold based on a characteristic amount of the liquid sample's scattered light.
Adaptive binarization using calculated threshold
The binarization unit binarizes the noise-added signals using a binarization threshold calculated by a binarization threshold calculation unit based on the liquid sample characteristics, rather than a fixed threshold.
Binarization threshold calculation based on RMS or DC values
The binarization threshold calculation unit computes the threshold from the RMS value or DC value of the scattered light component of the liquid sample, capturing measurement conditions tied to the sample.
Efficient generation of multiple uncorrelated noises from stored noise data
The device includes a storage unit for noise data of predetermined length, and a noise generation unit that produces the plural uncorrelated noises by taking different starting points in the stored noise data, thereby reducing memory requirements and facilitating parallel processing.
Calculation unit outputting average value of binarized signals
The calculation unit outputs an average value, such as a simple average, of the binarized signals resulting from parallel noise addition and binarization, to enhance signal detection.
The inventive features collectively enable improved particle detection in liquid samples by combining adaptive binarization thresholds calculated from sample characteristics with efficient stochastic resonance processing employing multiple uncorrelated noises generated from limited noise data, culminating in accurate detection of fine particles below existing measurable limits.
Stated Advantages
Enables detection of lower-level signals derived from particles by binarizing output signals using a threshold calculated based on RMS or DC values of the sample's scattered light component, thereby reducing misdetection.
Allows measurement of particles with smaller minimum measurable sizes than existing methods, improving detection accuracy and counting precision.
Automates calculation of the binarization threshold in conjunction with a reset operation, reducing time and effort for measurement preparation according to the sample.
Efficiently generates multiple uncorrelated noises from stored noise data by shifting starting points, reducing required noise data volume while supporting parallel processing.
Documented Applications
Measurement and counting of fine particles, including those less than 30 nm in size, suspended in liquid samples, particularly applied to semiconductor process monitoring where particle sizes smaller than 30 nm need to be detected.
Use in systems requiring precise liquid-borne particle detection with variable sample conditions such as refractive index, concentration, and temperature.
Interested in licensing this patent?