Azbil Corporation

Yamatake Corporation

Yamatake-honeywell Co., Ltd.

Tomoki Hosoi

Masashi Furuya

Daisuke Obara

Hirohiko Kazato

Seiichirou Kinugasa

Mitsuhiro Honda

A particle detecting device includes: a light measuring instrument that measures measured values for intensities of first, second, and third lights of mutually differing wavelengths, produced by particles to be measured; a boundary information storing portion that stores a non-linear discriminating boundary for separating a class of a first classification of particles and a class of a second classification of particles; and a particle classifying portion that classifies the particle being measured into either of the classifications for the first and second classifications of particles, based on measured values for the intensities of the first through third lights and on the discriminating boundary.

Particle detecting and discriminating device and method

A particle detecting device includes: a storing device that stores first boundary information wherein the third light intensity is recorded in a first range at a discriminating boundary for particles of first and second classifications, second boundary information wherein the third light intensity is recorded in a second range at a discriminating boundary for particles of first and second classifications, and discriminating information wherein identifiers for particles of the first and second classifications are recorded in cells bounded and not bounded by the discriminating boundary, respectively; and a particle identifying portion that evaluates a particle being measured as a particle of the first classification when the identifier for a particle of the first classification is acquired based on the measured values for the first and second light intensities and the measured value for the third light intensity falls between the first and second boundary values.

A detected temperature acquiring portion acquires detected temperatures from individual thermopile array sensors, a temperature change comparing portion compares changes in detected temperatures over time, for each respective location, between test locations selected from a overlapping region and individual comparing locations for two thermopile array sensors wherein planned detecting ranges partially overlap each other, to identify a best match location, a factor estimating portion, for each test location, generates an equation indicating a relationship between the best match location and positional shift factors, and establishes these equations as a system and solves the system of equations through a least-squares method to estimate the positional shift factors, and a detecting range identifying portion corrects location coordinates of the planned detecting ranges based on these positional shift factors to identify actual detecting ranges.

Device and method for identifying temperature detecting range

An image input section obtains a halftone image IG of a to-be-measured particle group. A binarization section binarizes the halftone image IT to obtain a binary image IB. A distance transform section performs distance transform for the binary image IB to obtain a distance-transformed image ID. In this case, the particle nucleus extraction section obtains particle nucleus candidates from the distance-transformed image ID and extracts a particle nucleus on the basis of the distance between the particle nucleus candidates and a separation parameter PS to obtain a particle nucleus image IC. A particle expansion section performs particle expansion processing for the particle nucleus image IC and distance-transformed image ID to obtain a particle-separated image IS. In this case, the particle expansion section performs particle expansion processing for the particle nucleus in the particle nucleus image IC along the value of distance transform to obtain the particle-separated image IS. A particle is recognized from the particle-separated image IS. In this way, a particle is recognized at a high speed and high accuracy.

Particle recognition device

According to a data conversion method and apparatus, an original signal is wavelet-transformed with a desired expansion count to generate transformed data. The transformed data is corrected such that a difference between the original signal and a value obtained by inversely wavelet-transforming the transformed data falls within a desired tolerance range, thereby generating the corrected transformed data as compressed data in which the original signal is compressed.

Growing community of inventors

Tokyo, Japan

Tomoki Hosoi