1.Home
Contact person
Eriko Watanabe, Professor
E-mail: ecma@fourier.ghrdp.uec.ac.jp
URL:http://www.fourier.lab.uec.ac.jp/
Department of Engineering Science
The University of Electro-Communications
URL: https://www.uec.ac.jp/eng/
ECMA-420
Device interface information for high-speed collation using holographic optical correlation
URL: ECMA-420 - Ecma International (ecma-international.org)
2.Technical information
2.1 Outline of holographic optical correlation system
Figure 1 shows the configuration of a typical holographic optical correlation system. The holographic optical correlation system consists of four units: a host computer unit, a data conversion unit, an optical correlation unit, and a judgment unit. The pink arrows in the diagram represent the creation of a hologram database, and the light blue arrows represent the process of matching images. The blue arrow labeled Device interface is the device Interface defined by ECMA-420 and represents the interface between the data processing/system controller unit and the holographic optical correlation unit. The functions of each unit are introduced below.
Figure 1 —— Configuration for holographic optical correlation system
(a) Host computer unit
This unit collects images for hologram database creation (TV, Internet, movies, packaged media, etc.), records holograms, and crawls images for matching. Then, upon receiving a collation request from the user PC, the hologram database data is searched and collated.
(b) Data conversion/system controller unit
Images for database creation are subjected to image size conversion, image feature extraction, and binarization with a constant white level. After that, it is converted into a pattern suitable for coaxial hologram recording (reference light is arranged in a ring shape on the outer area, object light is arranged in a circle shape inside area) and sent to the holographic light correlation unit via the device interface. On the other hand, the crawled image for collation is binarized by the same process as in the case of database creation, converted to become a circular object light, and sent to the holographic optical correlation unit. The system controller controls these series of operations as well as database creation and collation.
(c) Holographic optical correlation unit
Using the continuous processing image sent from the device interface, a holographic database is created (recorded). A pattern having a reference light part and an object light part on the left side of the optical correlation unit is used at the time of recording. Coaxial holograms with shift multiplexing technology records tens of millions of images on a single hologram disk. The recording speed of holograms is mainly limited by the display speed of the SLM and the laser power for recording. For this reason, holograms are recorded by rotating the disk at low speed. On the other hand, the crawled processed image for collation is a light pattern on the right side, and the holographic optical correlation is continuously performed by irradiating the readout light irradiated by this pattern to the recorded database hologram. The speed of the disk does not depend on the SLM or sensitivity of hologram materials. For this reason, holograms can be read out at a speed of several ten times faster than hologram recording. When the images match in the collation operation, a ring-shaped pattern corresponding to the reference light appears on the photodetector, and by detecting this, it can be detected whether or not there is a match.
(d) Judgment unit
The unit is a part of the optical correlation unit and determines whether the images match from the level of the obtained photodetector signal. The result is sent to the host PC via the device interface.
2.2 Additional information for Annex A (A.2), ECMA-420
Figure 2 shows the samples of detected optical correlated readout signal. Where holograms on the disk are recorded using the 1,000-frames movie stream of Japanese animation, which sampling rate is 10 frames/sec.
Figure 2-1 —— When No. 200 frame image is used as collation image, readout signal has a high peak signal at the hologram of No.200 frame. This figure is same as figure A.2, ECMA-420
Figure 2-2 —— When No. 400 frame image is used as collation image, readout signal has a high peak signal at the hologram of No.400 frame.
Figure 2-3 —— When No. 600 frame image is used as collation image, readout signal has a high peak signal at the hologram of No.600 frame.
Figure 2-4 —— When No. 800 frame image is used as collation image, readout signal has a high peak signal at the hologram of No.800 frame.
Figure 2 —— Samples of detected optical correlation signal
2.3 Reference
E. Watanabe, Y. Ichikawa, R. Akiyama, K. Kodate: Ultrahigh-Speed Optical Correlation System Using Holographic Disc, Jpn. J. Appl. Phys, 47, 5964-5967(2008)
K. Ikeda, E. Watanabe, "High-Speed Optical Correlator with Coaxial Holographic System", Jpn. J. Appl. Phys., 54, 09ME02(2015)
K. Ikeda, E. Watanabe, "High-speed image matching with coaxial holographic optical correlator", Jpn. J. Appl. Phys., 55, 09SC01 (2016)
K. Ikeda, H. Suzuki and E. Watanabe, "Optical Correlation-based Cross-domain Image Retrieval System", Optics Letters 42, 2603 (2017)
T. Hoshizawa, K. Saito, K. Ikeda, T. Sugaya and E. Watanabe, "Improvement of Correlation Speed of Holographic Optical Correlator by Low-Correlation Data Interleaving", Jpn. J. Appl. Phys., 58, SKKD06 (2019)
A. Inoue, R. Usami, K. Saito, Y. Honda, K. Ikeda and E. Watanabe, "Optical correlator-based computational ghost imaging towards high-speed computational ghost imaging", Jpn. J. Appl. Phys., 58, SKKA02 (2019)
K. Ikeda, A. Fukumoto, T. Sugaya, E. Watanabe, "Improving stability of coaxial holographic optical correlation system using a simple disk structure", Optical Review 26, 295 (2019)
https://www.mssf.or.jp/30fyhoukokusyo/30optical.pdf :Japanese, Contents of the project report
https://www.mssf.or.jp/30fyleaflet/30d2optical.pdf :Japanese, Outline of the project report
3.Intellectual property
Contact person: Eriko Watanabe, ecma@fourier.ghrdp.uec.ac.jp
Below is a list of the intellectual property we can provide:
◆Programs for image data transformation
(a) Feature extraction and white level control on binarization
@Binarization using edge detection
@Binarization using AI method
(b) Data pattern transformation for hologram database recording
(c) Data pattern transformation for collation
◆Know-how
Hologram recording technology, include hologram disk, equipment of hologram recording and read-out for optical correlation and high density hologram recording technology
◆Patent
JP 6851634 B2 2021.3.31
Name of Invention: Feature Transformation Module, Pattern Identification Device, Pattern Identification Method, and Program
4.Application for holographic optical correlation technology
4.1 General
Figure 3 shows an new application example of optical correlation technology, we will introduce research on so-called clairvoyance, which predicts the original object from light passing through a medium with fluctuations and/or scattering. In order to elucidate the scattering and fluctuation phenomena, we will relate the input light and the output light to the scattering and fluctuation medium, and try to create a database and model the scattering characteristics. At this time, correlation characteristics of the light field having various parameters such as amplitude, phase, wavelength, polarization, etc. are used. Here, focusing on whether the feature amount of the electromagnetic field (complex amplitude information of light waves, the inseparability of the polarization state and spatial distribution, etc.) is conserved before and after scattering, and to what extent it is conserved, a new index characterizing the scattering medium is identified, and it is used for the creation of a scattering characteristic database and the modeling of scattering / fluctuation field. In addition, we will propose and construct new methods for removing fluctuations and scattering by optical correlation imaging using a large-scale optical database, single-pixel imaging incorporating deep learning, and phase-shift digital holography. Furthermore, we aim to realize clairvoyance by trying scattering and fluctuation field modeling for biological tissues and optical communication domains targeting actual problems.
Figure 3 —— New application example using holographic optical correlation technology