DE102015007434A1 - information device - Google Patents

Abstract

The problem of established image recognition systems, which are to assign a record to a user-generated image of an object and display that data record to the user via a mobile computer, is that many available mathematical analysis methods make the identification safe, but a user-unacceptable processing time require. In the present case, the implementation of an essentially two-module image recognition is proposed which, despite reduced features, can offer the function which has been required for a long time surprisingly precisely and with virtually no delay in real time. Presentations such as exhibitions can be queried by means of common smartphones on the surprisingly simple application software by a user in individual objects by merely photographing the object on other, object-specific information. Especially with museums, fashion magazines and catalogs, this information device proved to be a virtually no perceptible delay working and versatile way to inform a user in addition and provide them with monetary, digital content. Through the rapid processing of the surprisingly effective module combination, both more targeted advertising and the direct online sale / sale of digital, related to the object products, better and more effectively possible. Print catalogs of real exhibitions become so obsolete and costs as well as environmental pollution are reduced considerably.

Description

TECHNICAL PART

The present invention relates to an information device according to the preamble of the independent claims. The technical field can generally be described as an area of digital object recognition. The computer-aided recognition of objects has long been the subject of research. So revealed the GB 972,054A as early as 1964 a system for pattern recognition. Such systems and methods can be used as in the GB 1208 126 A or the JP 54099529 A explained for the classification and / or identification of images / objects are used. The US 5,497,314 explains a corresponding use in a cashier system. Algorithms and methods that can be used to identify objects via their digital images are supplemental in the JP 01285809 A , of the US 5,245,675 A or even the JP 08007040 A explained.

DESCRIPTION OF THE PRIOR ART

Generic systems evaluate a digital image by means of a computer and, depending on the evaluation, provide additional object-related information involving a wireless communication network.

From the US 5,124,915 A For example, a computer-implemented method is known in which digital images are examined for patterns typical of fires or disasters in order to be able to automatically send corresponding messages to disaster services via radio signal. The disadvantage here is that fires are highly individual and extremely dynamic and therefore often elude conventional pattern recognition. Only by correlation with additional measured values such as gas composition or temperature are unique identifications possible. Notwithstanding these systems, the present invention is based on already known, presented objects, as they are also included in the following prior art.

From the US 6,023,241 A a multimedia player for exhibition visitors is known, which includes a computer with processor and memory and depending on a captured and processed via a CCD array code image provides additional data and / or provides a wirelessly connected exhibition server. From the DE 101 10 979 A1 a co-directional arrangement is known, which identifies pattern-specific data and wirelessly transmits / provides based on an optical pattern under detection and comparison with reference data. The disadvantage of these systems is that scanner patterns must be specified as metadata in order to be able to assign associated objects.

The opposite beats the DE 102 45 900 A1 a system in which a computer in the form of a smartphone is used, which downloads / provides additional and / or further information from a wireless network on the basis of digital object images of a real, known object.

A disadvantage of the generic method that no concrete modules or measures are specified, with which an image could be clearly assigned to a stored record. The technical domain has many freely available algorithms, but their interactions and possible combinations are unpredictable. In particular, for presentations of goods / services, the inventors of the prior art, no meaningful module packages are known in real time - so with virtually no perceptible delay - on a computer - preferably smart phone - running always and reliably offer a choice with the actually photographed object / actually photographed presentation would be identifiable.

The object of the present invention was therefore to overcome the disadvantages of the prior art and to provide a module package which, despite its speedy processing in real time on a computer with the performance of an average mobile computer - preferably an average smartphone - has its correct implementation and to provide meaningful identification of a photographed presentation.

The solution of this object is achieved according to the features of the independent claims. Advantageous embodiments will become apparent from the dependent claims and the following description.

SUMMARY OF THE INVENTION

According to the invention, the claimed information device is characterized in that providing at least the computer-implemented steps a) relative color analysis of the image; b01) identification of corners and angles within the image; b02) Identification of clearly visible corners and angles; b03) Correlation of optically clear corners and angles; c) combining the results from a) and b) in an image-specific amount of metadata; d) matching the image-specific metadata with a metadata database with identification of similar and identical, stored object images and creation of a hit list; e) providing the hit list for selecting a suitable object and / or reproducing object-specific information; includes.

DESCRIPTION OF THE INVENTION AND ADVANTAGEOUS CHARACTERISTICS

The generic information device comprises at least one mobile computer of at least one user and the mobile computer comprises at least one radio transceiver unit and a camera unit. The information device identifies object-related additional information by means of a user-created, digital image of the object. This happens with the cooperation of the radio transceiver unit. The object-related data are identified and provided. It is essential here that the provision of at least the computer-implemented steps a) relative color analysis of the image; b01) identification of corners and angles within the image; b02) Identification of clearly visible corners and angles; b03) Correlation of optically clear corners and angles; c) combining the results from a) and b) in an image-specific amount of metadata; d) matching the image-specific metadata with a metadata database with identification of similar and identical, stored object images and creation of a hit list; e) providing the hit list for selecting a suitable object and / or reproducing object-specific information; includes.

Preferably, the information device is characterized in that the step b03) the substeps b03-1) calculation of additional pixels with changed resolution; and b03-2) Enrichment of the pixels with information for the rotation correction; includes. Particularly preferred in this sub-step are distorted or non-differentiable by the observer angle point groups comprising corners and angles, checked for possible existence and calculated with high probability of presence and included in the subsequent processing as an additional feature.

Preferably, the information device is characterized in that the step d) the parallel substeps d01) feature comparison with stored object image metadata; and d02) feature comparison with the metadata of stored image sections; includes.

Preferably, the information device is characterized in that step d) comprises substeps d03) relative weighting of the hits; and / or d04) correlation with additional computer and / or user data; includes.

Preferably, the information device is characterized in that step d) the sub-step d05) correlation with at least one additional amount of data; comprising the additional data set selected from the group of partner preferences, commodity metadata, action metadata, campaign metadata, vendor metadata, digital object watermark, digital geo-watermark, location coordinates, object group, object maker, market metadata , Polling metadata, user profile metadata, metadata of other stored images, social network metadata, meta-data of Internet video portal, metadata of separate information systems preferred including newsletter or callback service, homepage metadata, raffle metadata, discount metadata, coupon metadata, temporal metadata of user behavior preferably includes time to purchase decision and / or activity profile, stock metadata, availability metadata, marketplace response metadata, user-feedback metadata, customer research metadata.

The information device is preferably characterized in that the object is at least one presentation object and is selected from the group consisting of museum presentation, shop-window presentation, poster presentation, magazine presentation, vehicle presentation, film presentation, dynamic product presentation , Advertising space presentation, product presentation, sales presentation, shelf presentation, display presentation, screen presentation, promotional display presentation.

Preferably, the information device is characterized in that the at least one mobile computer is selected from the group consisting of smartphone, tablet PC, PDA, laptop, camera, SLR camera, wearable.

The information device is preferably characterized in that the object is provided with at least one optical marking, which points to the additional, accessible information.

Preferably, the information device is characterized in that the optical tag is at least one tag selected from the group consisting of center tag, service provider tag, watermark tag, location tag, discount tag, geo cache tag.

Further advantages emerge from the exemplary embodiments. It is understood that the above-described, preferred features, advantages and subsequent embodiments are not restrictive. Advantageous or preferred, additional features and additional combinations of features, as described in the description, in the prior art and in the technical field, can be realized within the scope of the independent claims in the claimed subject-matter both individually and differently combined with would leave the invention.

BRIEF DESCRIPTION OF THE FIGURES

The figures illustrate on the basis of schematic diagrams in ...

1 the particularly advantageous data flow - a preferred embodiment - between a mobile computer, an image metadata database, an evaluation database and correlatable actions and research;

2 the processing chain of the image data of an advantageous embodiment up to the available hit list;

3 significant and advantageous processing steps of the digital image until a complete set of image metadata is obtained;

4 Cross correlation options of accessible data and selectable data in the final use of the result list.

DETAILED EXPLANATION OF THE INVENTION BY HAND OF EXAMPLES

• a) relative Farbanalyse des Bildes;
• b01) Identifikation von markanten Ecken und Winkeln innerhalb des Bildes;
• b02) Identifikation und Auswahl optisch deutlich erkennbarer, markanter Ecken und Winkel;
• b03) Korrelation der optisch deutlich erkennbaren Ecken und Winkel unter
• b03-1) Berechnung zusätzlicher Bildpunkte bei veränderter Auflösung; und
• b03-2) Anreicherung der Bildpunkte mit Informationen zur Rotationskorrektur;
• c) Zusammenführen der Ergebnisse aus a) und b) in einer bildspezifischen Menge an extrahierten Bild-Metadaten;
• d) Abgleich der bildspezifischen Metadaten mit einer Bild-Metadaten-Datenbank unter Identifikation ähnlicher und identischer, hinterlegter Objekt-Bilder unter Erstellung einer Trefferliste in Form einer Ergebnisliste unter
• d01) Merkmalsvergleich mit hinterlegten Objekt-Bild-Metadaten;
• d02-1) Merkmalsvergleich mit den Metadaten hinterlegter, bevorzugt überlappend angeordneter, Bild-Ausschnitte;
• d02-2) Erstellen einer Trefferliste mit zugehörigem Übereinstimmungsfaktor;
• d03-1) relative Gewichtung der Treffer unter Berücksichtigung zusätzlicher Geräte- und/oder Benutzer-Metadaten;
• d04-1) Korrelation mit zusätzlichen Metadaten, Computer- und/oder Benutzer-Daten ausgewählt aus der Gruppe bestehend aus Partnervorgaben, Benutzermetadaten, Gerätedaten, Verhaltensdaten, Auswertungsdaten, Kampagnendaten, Warendaten, Aktionsdaten, Benutzerdaten, Marktforschungsdaten, Meinungsforschungsdaten;
• d04-2) abschließendes Erstellen einer angepassten, gruppierten Ergebnisliste;
• e) Rückübertragen der Ergebnisliste auf ein Display des mobilen Computers und Bereitstellen der Trefferliste als Ergebnisliste zur Auswahl eines passenden Objekts und/oder Wiedergabe objektspezifischer Informationen;
• f01) Verknüpfung und/oder Weiterleitung des ausgewählten oder automatisch identifizierten Informationsdatensatzes mit mindestens einem weiteren Folge-Datensatz, der Folge-Datensatz ausgewählt aus der Gruppe bestehend aus Produktinformationen, Mediale Produktdarstellung, Online-Einkaufsmöglichkeit, Social-Media-Interaktion, Filialen in der Umgebung, Kampagnenaktion; und
• f02) Korrelation mit mindestens einem Service-Datensatz, der Service-Datensatz ausgewählt aus der Gruppe bestehend aus GPS-Position des mobilen Computers, Geräte-Daten, Benutzeraktionen, Kampagnenaktionen, Auswertungsergebnisse mindestens einer der vorbeschriebenen Datentypen, anonymisierte Kundenforschungs-Daten; umfasst sind.

In an advantageous embodiment, the information device comprising at least one mobile computer of at least one user, the mobile computer comprising a radio transceiver unit and a camera unit, wherein the information device object-related, additional information by means of a user-created, digital image of the object using the radio transceiver identifies and provides, characterized in that the providing comprises at least the computer-implemented analysis of the image by means of image algorithms for generating a result list, wherein in detail at least the steps

• a) relative color analysis of the image;
• b01) identification of prominent corners and angles within the image;
• b02) Identification and selection of clearly visible, prominent corners and angles;
• b03) Correlation of the visually distinct corners and angles below
• b03-1) calculation of additional pixels with changed resolution; and
• b03-2) Enrichment of the pixels with information for the rotation correction;
• c) combining the results of a) and b) in an image-specific amount of extracted image metadata;
• d) matching the image-specific metadata with an image metadata database with identification of similar and identical, stored object images under creation of a hit list in the form of a result list below
• d01) feature comparison with stored object image metadata;
• d02-1) feature comparison with the metadata stored, preferably overlapping arranged, image sections;
• d02-2) creating a hit list with associated match factor;
• d03-1) relative weighting of the hits taking into account additional device and / or user metadata;
• d04-1) correlation with additional metadata, computer and / or user data selected from the group consisting of partner preferences, user metadata, device data, behavioral data, evaluation data, campaign data, merchandise data, action data, user data, market research data, opinion research data;
• d04-2) finally creating a customized, grouped result list;
• e) retransferring the result list to a display of the mobile computer and providing the hit list as a result list for selecting a suitable object and / or reproducing object-specific information;
• f01) linking and / or forwarding the selected or automatically identified information record with at least one further sequence record, the sequence record selected from the group consisting of product information, media product presentation, online shopping, social media interaction, stores in the area , Campaign action; and
• f02) correlating with at least one service record, the service record selected from the group consisting of the GPS position of the mobile computer, device data, user actions, campaign actions, evaluation results of at least one of the above-described data types, anonymized customer research data; are included.

In a further, advantageous embodiment, a digital recording is created by a user with the aid of a smartphone camera. The smartphone is the mobile computer in this embodiment. The digitally generated image is subsequently in the smartphone and / or via a server available via the wireless transceiver unit in two aspects by means of image algorithms, in detail on a recognizable and derivable corners and angles and on the other hand on his color and Contrast distribution, examined.

By correlating the corners and angles, a rotation-corrected amount of image metadata is obtained, which is matched on the mobile computer and / or the server computer with a set of stored image metadata, creating a set of probable and highly probable hit images. During the comparison of individual aspects it is preferred to carry out first, relative tests on how characteristic the individual aspects of the photographed object are. With an already clear aspect - such as a unique, high-contrast black-and-white distribution - additional analysis can be dispensed with and an identification can be specified immediately. Otherwise a group of possible pictures will be put together. Images of lesser relevance are removed from the result list, the remaining images are divided into groups of different probability, and the thus prioritized and grouped result list is made available to the user in the form of characteristic pictograms for selection. By selecting one of the icons, the user can then open submenus, which retrieve him further, possible and / or available information and possibly complementary goods / services from the server computer and make it available for selection. Up to this point, no further computers or network pages were advantageously called. A use - preferably in a museum with a server installed in the museum - is thus independent of the Internet and can be done without the risk of virus contamination by external Internet addresses. In addition, such a use has the advantage that new exhibitions / exhibits can be detected in almost identical way in the system: It is only the permission of the mobile computer to change to capture the usual viewing and viewing angles by camera function and as a presentation-specific record in the Database. A further advantage of this method is that the object-specifically accessible information is centrally manageable: Current events or necessary corrections as well as adjustments to various exhibitions can be entered directly on the server computer of the museum and are from then on only in the corrected version retrievable. Particularly preferably, the software is provided as an application program via read-only radio networks at the entrance of the museum for established operating systems of smartphone users and allows users to use their own smartphone. For an additional charge, the saving of the retrieved information in the form of a user-definable tour with the additional, retrievable content is enabled and allowed by the software permanently for private, non-commercial use. In a first prototype, a server computer with Internet connection as described above was set up with hundreds of records of images of a fashion magazine and a standard smartphone with an average communication bandwidth was connected more than 10 km away via the Internet to the server. Subsequently, pictures of the fashion magazine as described above were recorded at extreme angles of inclination and sent to the server and the time until receipt of the hit list in the form of a pictogram. Group was measured. For comparison, QR codes from the same fashion magazine were recorded and called up and the time until the presentation of the website linked to the QR code was measured: image invention QR code (Time to result list) (Time to target page presentation) pullover 0.9 seconds 2.5 seconds jacket 0.6 seconds 2.7 seconds (high, typical contrast) shoe 1.1 seconds 3.2 seconds

In any case, the inventive processing of the image itself took only a few milliseconds in the analysis and the adjustment of the so specific reduced amount of data also takes place in the millisecond range. The essential time factor here is communication with the server / Internet, wherein the module combination according to the invention of the computer-implemented method always enables processing by one second + -0.3 seconds and therefore with virtually no perceptible delay - in real time. Particularly preferably, the software is supplemented by a module which first tests the quality / bandwidth of the radio connection, establishes a secure channel to the server computer and then informs the user whether and how a real-time use is possible; an advantageous possibility for acceleration may be the preparatory, location- and / or presentation-specific downloading of metadata and submenus, so that the mobile computer can offer the locally desired function almost independently of the server computer; The latter is particularly preferably linked to an updating function which preliminarily matches the possibly already stored data with the server computer. In a particularly advantageous embodiment, an exhibition or a museum or a presentation about cheap, rented to the visitors and individualized tablet PC's radio transceiver unit always highly up-to-date information highly, and optionally wirelessly connected to commercial recycling channels for corresponding goods or services , provide.

1 illustrates a particularly advantageous, fundamental data flow of an advantageous embodiment between a mobile computer, an image metadata database, an evaluation database and correlatable actions and researches. In detail, a mobile computer is reproduced here as a rectangle, the camera function is shown as a camera icon and includes at least one radio transceiver unit with; the illustrated as a thumbnail, user-created with the camera image and the feeding of its data in image algorithms can be done both in the mobile computer and in a wirelessly connected server computer. Depending on the performance of the mobile computer, one of both options may take place, most preferably both simultaneously. To graphically illustrate this selectable and preferably parallel processing of image data, image data and image algorithms are enclosed by a dashed extension of the module of the mobile computer. By correlating the thus accessible image metadata with an image metadata database, a result list is created, transmitted back to the mobile computer and presented to the user for selection and / or information. Pure information can take place if the result list has only one possible image with a high hit probability. Particularly advantageously, the user behavior via user metadata, device data and behavior data is recorded in parallel and stored in an evaluation database via evaluation algorithms; Depending on the release, the data in the database can optimize a results list, contribute additional, relevant results or provide links to pending, current campaigns, user actions or customer research. In a particularly preferred embodiment, a cooperation of different museums or even different shops is possible in which complementary presentations and offers a more comprehensive offer and more complete information and a discount system can support each other synergistically. In concrete detail so artists of a particular era and / or work or even complete product categories can be presented, informed about their local availability and possibly coupled with specific discounts on sightseeing / purchase the user made available.

2 illustrates a particularly advantageous processing chain of the image data of an advantageous embodiment to the available hit list in the form of a result list. In this case, the image metadata are extracted from the image data according to the invention and subjected to a feature comparison. The latter includes, on the one hand, the comparison of all features detected overall over the image area, but also distinctive features in certain quadrants positioned relative to the total area. Particularly preferred in this case in some image sections or object areas characteristic pattern - advantageously grid-like arranged contrast and / or color variations - identified; For example, it is possible to weave, dye or print textiles with specific grid patterns that allow one-to-one mapping. In addition, the implementation of location or position data is possible via such watermarks, which in turn allow conclusions about the scope and effectiveness of certain presentations. By the parallel checking and coupling of the totality of all features and features of individual image sections, the accuracy and / or speed of the evaluation algorithms of the feature comparison can advantageously be increased. The result is a hit list with match factors, which is subjected to a relative weighting of the hits. A correlation with additional device and / or user metadata allows a user-specific adaptation of the hit list: If, for example, certain artists, colors or even products have been excluded from the user, a corresponding message can be displayed here and the process can already be ended; Alternatively, similar and / or preferred products / hits can be increased in importance and also displayed as possible additional information. Both displays are possible at the same time. Finally, the adaptation of the hit list is correlated with partner specifications, whereby cooperation partners and / or the provider providing the information - preferably with consideration of the device and / or user metadata - adds or even eliminates additional selection options. The result list optimized in this way is displayed to the user and can also contain or offer links to other goods, campaigns or campaigns in the selection menu.

In a particularly advantageous embodiment according to 3 For example, a digital image in its data is subjected to significant and advantageous processing steps to provide a complete set of image metadata. In detail, a relative color analysis is carried out parallel to the one, which can also include a contrast analysis and allows conclusions about the color and / or contrast of the photographed object. On the other hand, an analysis of geometric elements, preferably corners and angles, made. Preferably, both analysis steps are carried out with increasingly increasing, particularly preferably in previously identified quadrants increasing, detail sharpness: If the recognizable in coarse resolution features still do not allow unambiguous identification, based on heretofore recognized features difference considerations are made, which the amount of meaningful and possible solutions already can restrict. In this way - starting from the first, recognized, prominent pixels or color components - quadrants or color values can be identified, which are still to be tested for a unique test with higher resolution in order to make this. Thus revealed 3 in addition to a continuous data flow and analysis control also a discontinuous, in the case of identified in coarse resolution, characterizing features, the subsequent resolutions set to constant and the analysis algorithms can be run without significant, further computational effort. By correlating directional vectors and angles to each other, geometrical features are considered to be linked in such a way that a rotation can no longer be distorted computationally: the geometric image features are therefore corrected with respect to their rotation; therefore one speaks also of the 'enrichment of the pixels with information for the rotation correction'; It is understood that in the case of implemented coordinate vectors, which can be given via position marks or underlaid, geometric patterns or prominent, horizontal or vertical lines in an object, a real rotation correction can take place - but mathematically is not absolutely necessary with sufficiently correlable features. In the illustrated processing, two parallel data streams are used, both of which analyze the image data and extract identifying metadata with increasing detail. Finally, the meaningfully extracted metadata is summarized in a file with image metadata and sent for further use. An essential element of this analysis is first the coupling of color distributions and / or contrast distributions with recognizable corners / angles. Preferably, this coupling is carried out in parallel steps, so that the detail sharpness in mutual analysis and review of the intermediate results is optimized and increased only until a sufficiently distinctive combination of image features is present. Particularly preferably, the analysis is carried out under control as a function of the probable and accessible hit suggestions, which can be preliminarily limited by supplementary additional information such as location, time or even on / in the object attached license plate.

In an advantageous combination with additional and / or parallel data fields, additional, monetary advantages can be achieved in accordance with the device claimed here 4 develop: Cross-correlation with accessible data and selectable data can be used in the final use of the results list meaningful statistics are generated, which would otherwise usually be determined by market or opinion research; By providing the individual user with various additional information, more concrete statements can be made:
Product information: Here you can determine which product information is of interest to the user and which are immediately rated as uninteresting with a click on 'next'. Furthermore, it can be determined which subordinate or external additional offers - if any - could be attractive to the user.
Media product presentation: Here you can determine which presentation effectively and effectively reaches and informs which user / user group; Through multimedia presentations, target group-optimized information can be distributed, which is of particular importance for recalls.
Online shopping: Here an identical, similar or derived product can be offered in parallel with the purchase; In a particularly preferred embodiment, for example, digital copies of the exhibited works including bonus material within the intranet of the exhibition are offered for sale in an exhibition.
Social Media Interaction: This provides the opportunity to disseminate important information about linked networks, to multiply the amount of informed users, and to get feedback on external review portals or rating mechanisms faster and more comprehensive.
Outlets in the area: This offers the possibility - in particular in combination with GPS data - to couple and display similar or identical presentations and / or available, commercially available goods with the interests of the user. Here, 'GPS data' in the sense of the present invention designates the data which determine the spatial position of the user and can advantageously be determined in a shopping center also by triangulation with the hotspots of the shopping center or with alternative satellite or positioning systems determine the geoposition / local position of the user. Consequently, GPS data is the local position data of a user, which can be read out or used directly or indirectly.
Campaign Action: This provides the ability to include short-lived discount or proposal systems; In a particularly advantageous embodiment, this is combined with digital watermarks stored in the presentations, whereby a task such as 'Find all different presentations of the object X in the city center' can already be evaluated digitally and rewarded depending on the speed of the compilation.
Device Data / User Action: Both the device data and the activities of the respective user can be meaningfully coupled with the previous aspects in order to filter and / or structure hit lists and result lists in a meaningful preparatory manner. For example, filtering of the contents is particularly preferably carried out in order to meaningfully avoid the presentation of results that would be unsuitable for a minor user in advance.
The encompassed, aforementioned aspects and also pending aspects can - preferably anonymously - via an evaluation database in customer research and / or user research incorporated in order to improve the information in general and target group-oriented.

In an advantageous embodiment, the inventive method is installed as an application program on a smartphone or similar, mobile computer. The proliferation of mobile computers (e.g., smartphones, tablets and smart watches, hereinafter referred to as "computers") provided with camera and transceiver equipment is steadily increasing. Almost every person is in permanent possession of a corresponding computer. People (hereinafter also referred to as "customers") today perceive media content and product presentations in a wide range of situations and places. If this information is printed, there is only a limited amount of room for further information.

Direct customer needs such as further product information, media (images, product videos, texts) or the opportunity to buy the presented product online or in a store in the vicinity, there is not. The customer must remember the ad when needed. For the advertiser presenting their products, there are limited and costly ways to conduct customer evaluations. Computer-mounted cameras and image recognition techniques developed in the present can be used to identify product presentations (for example, in newspapers, magazines, billboards, or shop windows), to enrich further information and customer interactions, and finally to draw conclusions about customer behavior. In image recognition, two essential steps are particularly advantageous. In the first, a full-surface image analysis is first performed to identify prominent pixels. This step is then carried out once again for individual quadrants. For this purpose, the image is divided into freely definable sections (for example, 3 × 3, 3 × 4, etc.). These sections are - preferably arranged overlapping - analyzed as an image to also calculate prominent pixels. Thanks to this, the customer does not have to take a picture of the whole picture; Computer-implemented can be an assignment of a Image section to a complete product presentation done. In the second step, relative color analyzes are carried out, which allow statements on image properties by means of individual color channels (for example red / green / blue or grayscale). The results of both analyzes are recorded in a result data set and compared with image data stored in databases. The resulting data and metadata in customer-specific databases are deposited particularly anonymously and, if the customer has given permission, are used for customer-specific support. In this way, customer behavior can be taken into account and analyzed with particular advantage. By using the described method, advertisers and business people can gain a wide variety of information: identifying and storing the location of the photographing process allows geographic conclusions - for example, the meaningfulness of advertising locations or regional differences in customer interest. Particularly advantageous here is a correlation with parallel device data of the mobile computer; Advantageously, a more concrete user profile can thus be created via click frequencies / history protocols / time allocations and frequency of use of the information service and / or GPS position data and used for the utilization and filtering of result lists and / or hit lists. Also, the absolute number of photographs of individual product presentations provides information about the interest of customers in the described object. So u. A. for presentations in magazines Derive the relative success of advertisements compared to any existing advertisements, or prove the value of individual advertising positions (for example, the greater value of advertisements on the first 20 pages of a magazine). Thousands of contact prices can also be determined more thoroughly. The collection of device metadata allows to collect information about the customer in relation to the hardware used. Thus statements are possible, which occupy a particularly high interest in a product (eg with Smartphone users of the manufacturer X). The added-value information and customer interactions provided in the context of the method described permit - also in relation to one another - conclusions about the motivation that causes a presentation of goods. For example, it can be measured more concretely whether the goods presentation of an automobile manufacturer leads to a particularly high rate of retrieval of the product video presented on the computer or the agreement of a test drive. All data collected at the individual user level can be fed to the entire database in order to gain further and cross-user inferences. In this particularly advantageous embodiment, the proposed information device offers both information for the user and relevant information from the field of opinion research and / or market research.

In an advantageous embodiment, the information device is correlated with action metadata of a discount system; In this case, the user is given a rough definition of the desired presentation / goods with the information that among all who find and photograph this presentation / goods in a certain period of time in a specific environment, a discount / price is distributed / raffled. Particularly preferably, the definition is given in the form of a riddle in order to additionally increase the incentive to solve.

In an advantageous embodiment, the information device is correlated with metadata for barrier-free presentations and thus offers the possibility of presenting, retrieving and selecting additional information and / or information purely via the audio or video channels of the mobile computer. A mobile computer without a display can thus be particularly preferred with headphones in areas in which loud play or viewing of animated sequences would distract from the actual presentation or the user are not possible.

In an advantageous embodiment, the information device is provided with a mobile computer whose camera unit also provides a film function, particularly preferably a film function depending on the viewing angle / viewing direction of the viewer. In this embodiment, the user can activate the movie function and watch a presentation, preferably a fashion show. The information device can then generate the corresponding metadata from the image data, identify hits and, depending on their duration of observation / frequency of consideration, group them into a hit list / result list. In this particularly advantageous embodiment, with the claimed device, a user may perceive / view a presentation with virtually no difficult or disturbing interaction. He just needs to turn the camera on and off at the beginning and end of the presentation. Particularly preferably, the switching on and off is effected by transmitting-receiving units at an input of the presentation, so that the user only deliberately activate the information device and at a suitable location - preferably by means of glasses with camera function and connection to the mobile computer - must wear. In any case, at the end of the presentation, a digitally stored overview will be available for the user and he will later be able to inform himself in detail about the individual, viewed objects / exhibits.

INDUSTRIAL APPLICABILITY

The problem of well-established image recognition systems, which use a mobile computer to associate a record with a user-created image of a real object and display that data set to the user, is that many mathematical analysis methods available in combination make the identification safe, but one user-side require unacceptable processing time. As a solution, the implementation of an essentially consisting of two modules image recognition is proposed, which despite reduced features surprisingly accurate and with almost no delay in real time can provide the long-needed function. Presentations such as exhibitions can be queried by means of common smartphones on the surprisingly simple application software by a user in individual objects by merely photographing the object on other, object-specific information. Especially with museums, fashion magazines and catalogs, this information device proved to be a virtually no perceptible delay working and versatile way to inform a user in addition and provide them with monetary, digital content. Through the rapid processing of the surprisingly effective module combination, both more targeted advertising and the direct online sale / sale of digital, related to the object products, better and more effectively possible. Print catalogs of real exhibitions become obsolete and costs as well as environmental pollution are reduced considerably.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• GB 972054 A [0001]
• GB 1208126 A [0001]
• JP 54099529 A [0001]
• US 5497314 [0001]
• JP 01285809 A [0001]
• US 5245675 A [0001]
• JP 08007040 A [0001]
• US 5124915 A [0003]
• US 6023241 A [0004]
• DE 10110979 A1 [0004]
• DE 10245900 A1 [0005]

Claims (10)

Information device comprising at least one mobile computer of at least one user, the mobile computer comprising a radio transceiver unit and a camera unit, the information device object-related, additional information by means of a user-created, digital image of the object in cooperation with the radio transceiver unit and providing, characterized in that providing at least the computer-implemented steps a) relative color analysis of the image; b01) identification of corners and angles within the image; b02) Identification of clearly visible corners and angles; b03) Correlation of optically clear corners and angles; c) combining the results from a) and b) in an image-specific amount of metadata; d) matching the image-specific metadata with a metadata database with identification of similar and identical, stored object images and creation of a hit list; e) providing the hit list for selecting a suitable object and / or reproducing object-specific information; includes. Information device according to the preceding claim, characterized in that the step b03) the substeps b03-1) calculation of additional pixels with changed resolution; and b03-2) Enrichment of the pixels with information for the rotation correction; includes. Information device according to one of the preceding claims, characterized in that the step d) the parallel sub-steps d01) feature comparison with stored object image metadata; and d02) feature comparison with the metadata of stored image sections; includes. Information device according to one of the preceding claims, characterized in that step d) the substeps d03) relative weighting of the hits; and / or d04) correlation with additional computer and / or user data; includes. Information device according to one of the preceding claims, characterized in that step d) the sub-step d05) correlation with at least one additional amount of data; comprising the additional data set selected from the group of partner preferences, commodity metadata, action metadata, campaign metadata, vendor metadata, digital object watermark, digital geo-watermark, location coordinates, object group, object maker, market metadata , Polling metadata, user profile metadata, metadata of other stored images, social network metadata, meta-data of Internet video portal, metadata of separate information systems preferred including newsletter or callback service, homepage metadata, raffle metadata, discount metadata, coupon metadata, temporal metadata of user behavior preferably including time to purchase decision and / or activity profile, stock metadata, availability metadata, marketplace response metadata, user-feedback metadata, customer research metadata. Information device according to one of the preceding claims, characterized in that the object is at least one presentation object and is selected from the group consisting of museum presentation, shop window presentation, poster presentation, magazine presentation, vehicle presentation, film presentation , dynamic merchandise presentation, advertising space presentation, product presentation, sales presentation, shelf presentation, display presentation, screen presentation, promotional display presentation. Information device according to one of the preceding claims, characterized in that the at least one mobile computer is selected from the group consisting of smartphone, tablet PC, PDA, laptop, camera, SLR camera, wearable. Information device according to one of the preceding claims, characterized in that the object is provided with at least one optical marking, which points to the additional, accessible information. An information device according to the preceding claim, characterized in that the optical tag is at least one tag selected from the group consisting of center tag, service provider tag, watermark tag, location tag, discount tag, geo cache tag. Information device, according to one of the preceding claims, comprising at least one mobile computer of at least one user, the mobile computer comprising a radio transceiver unit and a camera unit, wherein the information device object-related, additional Identifying and providing information by means of a user-generated digital image of the object with the aid of the radio transceiver unit, characterized in that the providing comprises at least the computer-implemented analysis of the image by means of image algorithms for generating a result list, wherein at least the steps of a) relative color analysis of the image; b01) identification of prominent corners and angles within the image; b02) Identification and selection of clearly visible, prominent corners and angles; b03) Correlation of optically clear corners and angles under b03-1) Calculation of additional pixels with changed resolution; and b03-2) Enrichment of the pixels with information for the rotation correction; c) combining the results of a) and b) in an image-specific amount of extracted image metadata; d) matching the image-specific metadata with an image metadata database with identification of similar and identical, stored object images and creating a hit list in the form of a result list under d01) feature comparison with stored object image metadata; d02-1) feature comparison with the metadata of stored - preferably overlapping arranged - image sections; d02-2) creating a hit list with associated match factor; d03-1) relative weighting of the hits taking into account additional device and / or user metadata; d04-1) correlation with additional metadata, computer and / or user data selected from the group consisting of partner preferences, user metadata, device data, behavioral data, evaluation data, campaign data, merchandise data, action data, user data, market research data, opinion research data; d04-2) finally creating a customized, grouped result list; e) retransferring the result list to a display of the mobile computer and providing the hit list as a result list for selecting a suitable object and / or reproducing object-specific information; f01) linking and / or forwarding the selected or automatically identified information record with at least one further sequence record, the sequence record selected from the group consisting of product information, media product presentation, online shopping, social media interaction, stores in the area , Campaign action; and f02) correlating with at least one service record, the service record selected from the group consisting of GPS position of the mobile computer, device data, user actions, campaign actions, evaluation results of at least one of the above-described data types, anonymized customer research data; are included.

Images