{"id":36741,"date":"2019-10-31T22:13:29","date_gmt":"2019-10-31T19:13:29","guid":{"rendered":"https:\/\/prohoster.info\/blog\/uvidet-pochti-nevidimoe-eshhe-i-v-tsvete-metodika-vizualizatsii-obektov-cherez-rasseivatel\/"},"modified":"2019-10-31T22:13:29","modified_gmt":"2019-10-31T19:13:29","slug":"uvidet-pochti-nevidimoe-eshhe-i-v-tsvete-metodika-vizualizatsii-obektov-cherez-rasseivatel","status":"publish","type":"post","link":"https:\/\/prohoster.info\/en\/blog\/news\/uvidet-pochti-nevidimoe-eshhe-i-v-tsvete-metodika-vizualizatsii-obektov-cherez-rasseivatel","title":{"rendered":"See the nearly invisible, now in color: a methodology for visualizing objects through a scatterer","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"

\"See
\n
\nOne of Superman's most famous abilities is his super vision, which allowed him to examine atoms, see in the dark and at great distances, and even see through objects. This ability is rarely demonstrated on screen, but it exists. In our reality, it is also possible to see through nearly opaque objects by applying certain scientific tricks. However, the resulting images have always been black and white until recently. Today we will explore a study in which researchers from Duke University (USA) managed to capture a colored image of objects hidden behind an opaque wall using a single light flash. What is this super technology, how does it work, and in what areas can it be applied? We will learn about this from a report by the research group. Let's go.<\/a><\/noindex><\/p>\n

The foundation of the research<\/h3>\n

\nDespite all the potential 'perks' of the object visualization technology in scattering environments, there are a number of implementation issues with this technology. The main problem is that the paths of photons passing through the scatterer vary greatly, leading to random patterns speckles*<\/i> on the other side.<\/p>\n

\"See
\nSpeckle*<\/b> is a random interference pattern formed by the mutual interference of coherent waves, which have random phase shifts and\/or a random set of intensities. It usually appears as a collection of bright spots (dots) on a dark background.<\/p><\/blockquote>\n

In recent years, several visualization methods have been developed that allow bypassing scatterer effects and extracting information about the object from the speckle pattern. The problem with these methods is their limitations \u2014 one needs to have certain knowledge about the object, access to the scattering medium or the object itself, etc.<\/p>\n

At the same time, there exists a more advanced method, according to scientists \u2014 memory effect visualization (ME). This method allows for visualizing an object without prior knowledge concerning the object itself or the scattering medium. All methods have their drawbacks, and the ME method is no exception. To obtain high-contrast speckle patterns and, respectively, more accurate images, the illumination must be narrowband, i.e., less than 1 nm.<\/p>\n

It is also possible to outsmart the limitations of the ME method, but again, these tricks are related to access to the optical source or object before the scatterer, or with direct measurement. PSF*<\/i>.<\/p>\n

PSF*<\/b> \u2014 point spread function, describing the image that the imaging system obtains when observing a point light source or point object. <\/p><\/blockquote>\n

Researchers consider these methods to be functional but not perfect, as measuring the PSF is not always possible due to, for example, the dynamic nature of the scatterer or its inaccessibility prior to the visualization procedure. In other words, there is work to be done. <\/p>\n

In their work, the researchers propose an alternative approach. They demonstrate a method for implementing multispectral visualization of objects through a scattering medium using a single speckle measurement with a monochrome camera. Unlike other techniques, this one does not require prior knowledge of the PSF system or the source spectrum.<\/p>\n

The new method allows for the creation of high-quality images of the target object in five well-separated spectral channels between 450 nm and 750 nm, which has been confirmed through calculations. In practice, however, it has so far been possible to achieve visualization of three well-separated spectral channels between 450 nm and 650 nm and six adjacent spectral channels between 515 and 575 nm.<\/p>\n

Principle of the new method<\/h3>\n

\n\"See
\nImage #1: lamp \u2014 spatial light modulator \u2014 scatterer (with an iris diaphragm) \u2014 encoding aperture \u2014 prism \u2014 optical relay (1:1 visualization) \u2014 monochrome camera.<\/i><\/p>\n

Researchers note three main elements of any visualization through a scatterer: the object of interest (illuminated externally or self-luminous), the scatterer, and the detector.<\/p>\n

As in standard ME systems, this study examines an object whose angular size lies within the field of view of the ME and is at a distance u behind the scatterer. After interacting with the scatterer, light travels a distance v before reaching the detector. <\/p>\n

Conventional ME visualization employs standard cameras, while this method utilizes a module of the encoding detector made up of an encoding aperture and an optical element dependent on the wavelength. The purpose of this element is to uniquely modulate each spectral channel before their combination and conversion in a monochrome detector.<\/p>\n

Thus, instead of merely measuring a low-contrast speckle, whose spectral channels are inseparably mixed, a spectrally multiplexed signal was recorded, which is well suited for separation.<\/p>\n

The researchers emphasize once again that their method does not require any a priori known characteristics or assumptions about the scatterer or light source.<\/p>\n

After preliminary measurements of the multiplexed speckle, a known value of T\u03bb (a wavelength-dependent encoding pattern) was used for individual reconstruction of the speckle in each spectral band.<\/p>\n

In their work at the calculation and modeling stage, the scientists applied certain machine learning methods that can assist in implementing a previously unaddressed method. Primarily, sparse matrix feature learning was used to represent the speckle. <\/p>\n

Feature Learning*<\/b> \u2014 allows the system to automatically find the representations necessary for identifying the features of the original data.<\/p><\/blockquote>\n

As a result, a database was generated, trained on speckle images from various measurement configurations. This database is sufficiently generalized and does not depend on specific objects and scatterers involved in generating the mask I\u03bbx, y. In other words, the system learns based on a scatterer that is not used in the experimental configuration, i.e., the system does not have access to it, as the researchers intended.<\/p>\n

To obtain speckle images at each wavelength, the OMP algorithm was used (orthogonal matching pursuit<\/i>).<\/p>\n

Ultimately, by independently calculating the autocorrelation of each spectral channel and inverting the autocorrelation at each wavelength, images of the object were obtained. The images obtained at each wavelength are then combined to create a color image of the object.<\/p>\n

\"See
\nImage #2: step-by-step process of constructing an image of the object.
\nAccording to its creators, this method makes no assumptions about correlations between spectral channels and only requires the assumption that the wavelength value is sufficiently random. Additionally, this method only requires information about the encoding detector, relying on the pre-calibration of the encoding aperture and a pre-trained data library. These characteristics make this visualization method quite versatile and non-invasive. <\/p>\n

Simulation Results<\/h3>\n

\nTo begin with, let's consider the results of the simulation. <\/p>\n

\"See
\nImage #3<\/i><\/p>\n

The image above shows examples of multispectral images of two objects taken through a scatterer. The top row in 3a<\/b> contains the object of interest, consisting of several numbers displayed in both false color and in breakdown by spectral channel. When constructing the object in false color, the intensity profile of each wavelength is represented in CIE 1931 RGB space.<\/p>\n

The reconstructed object (bottom row in 3a<\/b>) in both false color and in terms of individual spectral channels demonstrates that the methodology provides excellent visualization with only minimal cross-talk between spectral channels, which does not play a significant role in the process.<\/p>\n

After obtaining the reconstructed object, i.e. after visualization, it was necessary to assess the degree of accuracy by comparing the spectral intensity (averaged over all bright pixels) of the actual object and the reconstructed one (3b<\/b>).<\/p>\n

In the images 3c<\/b> the actual object (top row) and the reconstructed image (bottom row) for a cotton stem cell are shown, while in 3d<\/b> the accuracy of the visualization analysis is presented.<\/p>\n

To assess the accuracy of the visualization, it was necessary to calculate the values of the Structural Similarity Index (SSIM) and the Peak Signal-to-Noise Ratio (pSNR) of the actual object for each spectral channel.<\/p>\n

\"See
\n
\nThe table above shows that each of the five channels has an SSIM value of 0.8\u20130.9 and a PSNR of over 20. This suggests that despite the low contrast of the speckle signal, overlaying the detector with five spectral bands of 10 nm width allows for a sufficiently accurate reconstruction of the spatial-spectral properties of the studied object. In other words, the methodology works, but these are just simulation results. To ensure their findings, the researchers conducted a series of practical experiments.<\/p>\n

Results of the experiments<\/h3>\n

\nOne of the most significant differences between simulation and real experiments is the environment, i.e., the conditions under which both are conducted. In the former case, controlled conditions are present; in the latter, unpredictable conditions apply.<\/p>\n

Three spectral channels with widths of 8-12 nm centered at 450, 550, and 650 nm were examined, which combined with various relative magnitudes generate a wide range of colors.<\/p>\n

\"See
\nImage No. 4<\/i><\/p>\n

The image above shows a comparison between the actual object (the multicolored letter 'H') and the reconstructed one. The light exposure time (shutter speed, i.e., exposure) was set to 1800 s, allowing for an SNR of around 60-70 dB. According to the researchers, this SNR value is not critical for the experiment but serves as additional confirmation of the effectiveness of their method, especially with complex objects. In reality, outside of laboratory conditions, this method can be significantly faster.<\/p>\n

In the top row of Image No. 4, the object is shown at each wavelength (from left to right) along with the actual full-color object.<\/p>\n

To obtain an image of the real object during visualization, a machine vision camera was used with appropriate bandpass filters to directly display the spectral components and achieve a full-color image by summing the resulting spectral channels.<\/p>\n

The second row of the image above shows the autocorrelation patterns of each reconstructed spectral channel, forming multiplexed measurements that serve as input for the data processing stage.<\/p>\n

The third row represents the reconstructed object in each spectral channel, as well as the reconstructed full-color object, i.e., the final visualization result.<\/p>\n

The full-color image indicates that the relative magnitudes between the spectral channels are also correct, as the color of the combined reconstructed image corresponds to the actual value, and the SSIM coefficient exceeds 0.92 for each channel.<\/p>\n

The bottom row confirms this statement by demonstrating the comparison of the intensity of the real object and the reconstructed one. The data of both coincide across all spectral ranges.<\/p>\n

This indicates that even the presence of noise and potential modeling errors did not prevent the acquisition of a high-quality image, and the experimental results correlate well with the modeling outcomes.<\/p>\n

The aforementioned experiment was conducted with separate spectral channels. Scientists conducted another experiment, but this time with adjacent channels, specifically with a continuous spectral range of 60 nm.<\/p>\n

\"See
\nImage No. 5<\/i><\/p>\n

As the real object, the letter 'X' and the '+' sign were used (5a<\/b>). The spectrum of the letter 'X' is relatively uniform and continuous \u2014 between 515 and 575 nm, whereas the '+' has a structured spectrum, predominantly located between 535 and 575 nm (5b<\/b>). For this experiment, an exposure of 120 seconds was used to achieve the desired (as before) SNR of 70 dB.<\/p>\n

A bandpass filter of 60 nm width was also applied over the entire object, along with a low-pass filter over the '+' sign. During reconstruction, the 60 nm spectrum is divided into 6 adjacent channels of 10 nm width (5b<\/b>).<\/p>\n

As we can see in the images 5c<\/b>, the resulting images align excellently with the real object. This experiment showed that the presence or absence of spectral correlations in the measured speckle does not affect the effectiveness of the studied visualization technique. The researchers believe that a far greater role in the visualization process\u2014specifically in its success\u2014plays not so much the spectral characteristics of the object, but rather the calibration of the system and the details of its encoding detector.<\/p>\n

Epilogue<\/h3>\n

\nIn this work, the scientists describe a new method of multispectral imaging through a scatterer. The modulation of the speckle, dependent on the wavelength, using a coding aperture allowed for a single multiplexed measurement and the calculation of speckle using the OMP algorithm based on machine learning. <\/p>\n

Using the example of a multi-colored letter 'N', the researchers showed that focusing on five spectral channels corresponding to violet, green, and three shades of red results in an image reconstruction that contains all colors of the original (blue, yellow, etc.). <\/p>\n

According to the researchers, their technique could be beneficial in both medicine and astronomy. Color carries important information in both fields: in astronomy\u2014 the chemical composition of the studied objects, in medicine\u2014 the molecular composition of cells and tissues.<\/p>\n

At this stage, the scientists note only one problem that may cause inaccuracies in visualization, which is modeling errors. Due to the considerable time required to carry out the process, changes in the environment may occur that introduce their own corrections, not accounted for during preparation. However, it is planned to find a way to mitigate this issue in the future, allowing the described visualization technique to be not only accurate but also stable under any conditions.<\/p>\n

Friday off-topic:<\/b>

\r\n \"Play\r\n
<\/div>\r\n <\/div><\/center>
\nLight, color, music, and the trio of the world's most famous blue 'freaks' (Blue Man Group).<\/i><\/p>\n

Thank you for your attention, stay curious, and have a great weekend, everyone! \ud83d\ude42<\/p>\n

\nThank you for staying with us. Do you enjoy our articles? Would you like to see more interesting materials? Support us by placing an order or recommending us to your friends. 30% discount for Habr users on a unique entry-level server designed by us for you:<\/b> The whole truth about VPS (KVM) E5-2650 v4 (6 Cores) 10GB DDR4 240GB SSD 1Gbps starting at $20, or how to properly divide a server?<\/a><\/noindex> (options available with RAID1 and RAID10, up to 24 cores and up to 40GB DDR4).<\/p>\n

Dell R730xd for half the price?<\/b> Only with us 2 x Intel TetraDeca-Core Xeon 2x E5-2697v3 2.6GHz 14C 64GB DDR4 4x960GB SSD 1Gbps 100TB starting at $199<\/a><\/noindex> in the Netherlands! Dell R420 \u2014 2x E5-2430 2.2GHz 6C 128GB DDR3 2x960GB SSD 1Gbps 100TB \u2014 from $99!<\/b><\/b> Read about how To build a corporate-class infrastructure using Dell R730xd E5-2650 v4 servers costing 9000 euros for peanuts?<\/a><\/noindex>
\n
Source: habr.com<\/a><\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"

\u041e\u0434\u043d\u043e\u0439 \u0438\u0437 \u0441\u0430\u043c\u044b\u0445 \u0437\u043d\u0430\u043c\u0435\u043d\u0438\u0442\u044b\u0445 \u0441\u043f\u043e\u0441\u043e\u0431\u043d\u043e\u0441\u0442\u0435\u0439 \u0421\u0443\u043f\u0435\u0440\u043c\u0435\u043d\u0430 \u044f\u0432\u043b\u044f\u0435\u0442\u0441\u044f \u0441\u0443\u043f\u0435\u0440\u0437\u0440\u0435\u043d\u0438\u0435, \u043a\u043e\u0442\u043e\u0440\u043e\u0435 \u043f\u043e\u0437\u0432\u043e\u043b\u044f\u043b\u043e \u0435\u043c\u0443 \u0440\u0430\u0441\u0441\u043c\u0430\u0442\u0440\u0438\u0432\u0430\u0442\u044c \u0430\u0442\u043e\u043c\u044b, \u0432\u0438\u0434\u0435\u0442\u044c \u0432 \u0442\u0435\u043c\u043d\u043e\u0442\u0435 \u0438 \u043d\u0430 \u043e\u0433\u0440\u043e\u043c\u043d\u043e\u0435 \u0440\u0430\u0441\u0441\u0442\u043e\u044f\u043d\u0438\u0435, \u0430 \u0435\u0449\u0435 \u0432\u0438\u0434\u0435\u0442\u044c \u0441\u043a\u0432\u043e\u0437\u044c \u043f\u0440\u0435\u0434\u043c\u0435\u0442\u044b. \u042d\u0442\u0443 \u0441\u043f\u043e\u0441\u043e\u0431\u043d\u043e\u0441\u0442\u044c \u043a\u0440\u0430\u0439\u043d\u0435 \u0440\u0435\u0434\u043a\u043e \u0434\u0435\u043c\u043e\u043d\u0441\u0442\u0440\u0438\u0440\u0443\u044e\u0442 \u043d\u0430 \u044d\u043a\u0440\u0430\u043d\u0430\u0445, \u043d\u043e \u043e\u043d\u0430 \u0435\u0441\u0442\u044c. \u0412 \u043d\u0430\u0448\u0435\u0439 \u0436\u0435 \u0440\u0435\u0430\u043b\u044c\u043d\u043e\u0441\u0442\u0438 \u0432\u0438\u0434\u0435\u0442\u044c \u0441\u043a\u0432\u043e\u0437\u044c \u043f\u0440\u0430\u043a\u0442\u0438\u0447\u0435\u0441\u043a\u0438 \u043f\u043e\u043b\u043d\u043e\u0441\u0442\u044c\u044e \u043d\u0435\u043f\u0440\u043e\u0437\u0440\u0430\u0447\u043d\u044b\u0435 \u043e\u0431\u044a\u0435\u043a\u0442\u044b \u0442\u0430\u043a\u0436\u0435 \u043c\u043e\u0436\u043d\u043e, \u043f\u0440\u0438\u043c\u0435\u043d\u0438\u0432 \u043d\u0435\u043a\u043e\u0442\u043e\u0440\u044b\u0435 \u043d\u0430\u0443\u0447\u043d\u044b\u0435 \u0442\u0440\u044e\u043a\u0438. \u041e\u0434\u043d\u0430\u043a\u043e, \u043f\u043e\u043b\u0443\u0447\u0435\u043d\u043d\u044b\u0435 \u0441\u043d\u0438\u043c\u043a\u0438 \u0432\u0441\u0435\u0433\u0434\u0430 […]<\/p>\n","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"author":1,"featured_media":27522,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[702],"tags":[],"class_list":["post-36741","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\n","aioseo_head_json":{"title":"\ud83e\udd47See the almost invisible, and even in color: a method for visualizing objects through a diffuser | ProHoster","description":".","canonical_url":"https:\/\/prohoster.info\/en\/blog\/news\/uvidet-pochti-nevidimoe-eshhe-i-v-tsvete-metodika-vizualizatsii-obektov-cherez-rasseivatel","robots":"max-image-preview:large","keywords":"","webmasterTools":{"miscellaneous":""},"schema":null,"og:locale":"en_US","og:site_name":"ProHoster | \u041a\u0443\u043f\u0438\u0442\u044c \u043d\u0430\u0434\u0435\u0436\u043d\u044b\u0439 \u0445\u043e\u0441\u0442\u0438\u043d\u0433 \u0434\u043b\u044f \u0441\u0430\u0439\u0442\u043e\u0432 \u0441 \u0437\u0430\u0449\u0438\u0442\u043e\u0439 \u043e\u0442 DDoS, VPS VDS \u0441\u0435\u0440\u0432\u0435\u0440\u044b","og:type":"article","og:title":"\ud83e\udd47\u0423\u0432\u0438\u0434\u0435\u0442\u044c \u043f\u043e\u0447\u0442\u0438 \u043d\u0435\u0432\u0438\u0434\u0438\u043c\u043e\u0435, \u0435\u0449\u0435 \u0438 \u0432 \u0446\u0432\u0435\u0442\u0435: \u043c\u0435\u0442\u043e\u0434\u0438\u043a\u0430 \u0432\u0438\u0437\u0443\u0430\u043b\u0438\u0437\u0430\u0446\u0438\u0438 \u043e\u0431\u044a\u0435\u043a\u0442\u043e\u0432 \u0447\u0435\u0440\u0435\u0437 \u0440\u0430\u0441\u0441\u0435\u0438\u0432\u0430\u0442\u0435\u043b\u044c | ProHoster","og:description":".","og:url":"https:\/\/prohoster.info\/en\/blog\/news\/uvidet-pochti-nevidimoe-eshhe-i-v-tsvete-metodika-vizualizatsii-obektov-cherez-rasseivatel","og:image":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:secure_url":"https:\/\/prohoster.info\/wp-content\/uploads\/2021\/11\/logo-350.jpg","og:image:width":350,"og:image:height":350,"article:published_time":"2019-10-31T19:13:29+00:00","article:modified_time":"2019-10-31T19:13:29+00:00","article:publisher":"https:\/\/www.facebook.com\/prohoster","article:author":"https:\/\/www.facebook.com\/prohoster"},"aioseo_meta_data":{"post_id":"36741","title":null,"description":null,"keywords":null,"keyphrases":null,"primary_term":null,"canonical_url":null,"og_title":null,"og_description":null,"og_object_type":"default","og_image_type":"default","og_image_url":null,"og_image_width":null,"og_image_height":null,"og_image_custom_url":null,"og_image_custom_fields":null,"og_video":null,"og_custom_url":null,"og_article_section":null,"og_article_tags":null,"twitter_use_og":false,"twitter_card":"default","twitter_image_type":"default","twitter_image_url":null,"twitter_image_custom_url":null,"twitter_image_custom_fields":null,"twitter_title":null,"twitter_description":null,"schema":{"blockGraphs":[],"customGraphs":[],"default":{"data":{"Article":[],"Course":[],"Dataset":[],"FAQPage":[],"Movie":[],"Person":[],"Product":[],"ProductReview":[],"Car":[],"Recipe":[],"Service":[],"SoftwareApplication":[],"WebPage":[]},"graphName":"","isEnabled":true},"graphs":[]},"schema_type":null,"schema_type_options":null,"pillar_content":false,"robots_default":true,"robots_noindex":false,"robots_noarchive":false,"robots_nosnippet":false,"robots_nofollow":false,"robots_noimageindex":false,"robots_noodp":false,"robots_notranslate":false,"robots_max_snippet":null,"robots_max_videopreview":null,"robots_max_imagepreview":"large","priority":null,"frequency":null,"local_seo":null,"seo_analyzer_scan_date":"2026-01-22 04:40:22","breadcrumb_settings":null,"limit_modified_date":false,"reviewed_by":null,"ai":null,"created":"2021-02-28 11:13:39","updated":"2026-01-22 04:40:22","focus_keyword":null,"additional_keywords":null,"truseo_locale":null},"gt_translate_keys":[{"key":"link","format":"url"}],"_links":{"self":[{"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/posts\/36741","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/comments?post=36741"}],"version-history":[{"count":0,"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/posts\/36741\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/media\/27522"}],"wp:attachment":[{"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/media?parent=36741"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/categories?post=36741"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/prohoster.info\/en\/wp-json\/wp\/v2\/tags?post=36741"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}

For a more detailed understanding of the nuances of the study, I recommend checking out the scientists' report<\/a><\/noindex> and additional materials<\/a><\/noindex> related to it.<\/p>\n