FR3063170B1 - DEVICE FOR RECORDING INFORMATION FROM A CAPTURE OF AN IMAGE OF A DOCUMENT - Google Patents

Abstract

The invention relates to a device for recording information from a capture of an image of a document, comprising: a terminal provided with an objective of image capture, a screen and means telecommunication device, - a support for holding the terminal on a support surface, in a predetermined position in which its screen points above the horizontal, - a reflection module mounted on the support provided with a reflecting surface disposed opposite the objective of the terminal so as to form when the terminal is held by the support, an image of an area of the support surface located on one side of the support and viewable by the capture lens of image of the terminal, the terminal being provided with means for correcting the dimensions and perspective of the image of an object of regular shape disposed in said zone to best match the dimensions and the appearance of said object.

Description

DEVICE FOR RECORDING INFORMATION FROM A CAPTURE OF AN IMAGE OF A DOCUMENT

Field of the invention

The present invention relates to the field of recording information from a capture of an image of a document.

More particularly, the document in question may be an A4 fomat page for example, and the information to be recorded, consist of alphanumeric and / or graphic information. The invention finds a particular application in the context of the management of human resources information, retrieved for example from a call for applications for a position to power, and on the occasion of which a recruiter receives tens, hundreds thousands of applications in the form of cover letters and resumes, this reception being done over a relatively long period for applications in response to an offer of post, or in a very limited time for applications received during a professional forum.

It is then a question for the recruiter to gather a maximum of information on each of the candidates in a format allowing him an objective and relevant comparison of these in a time as small as possible, and a recording on the duration of these data to recontact candidates that may be relevant to a current or future announcement, and schedule reminders.

State of the art and its disadvantages

The tools for document scanning and character recognition and / or existing available images certainly allow such records of information and the constitution of a reliable database but at the cost of a long and tedious treatment and susceptible to involve multiple collaborators to scan the document, retrieve and enter data from this document and add comments from the recruiter who eventually conducted a professional interview with the candidate.

More global solutions have emerged.

For example, it is now known, in particular from EP 0533544, a method and device for processing alphanumeric and graphical information for constituting a database.

According to the teaching of this document, the method processes alphanumeric and graphical information recorded on a medium in the form of pages, to form a searchable databank and to obtain the editing of information.

According to this method, the following steps are automatically performed: A) determination of digitized image pages; B) a first phase of processing the digitized image pages allowing: the verification that these pages are sequential, the determination of the characteristic elements of each page, the determination of the angle of deflection, the manufacture of a mask, a windowing, the rectification of alignment, identification of characteristic elements; Segmentation of images recording on separate files of scanned image pages, scanned image entities, digitized segmented images that can be edited; • C) a second phase of processing consisting in performing an optical character recognition of the alphanumeric information blocks from the segmented images and recording the data on a file constituting a searchable databank.

If the teaching of this document appears satisfactory in terms of retrieving relevant data from a document and storing it at a specific location in a database, providing only one person, there is no no less tacit as to the influence that the position and / or inclination of the document under consideration may have on the scanning tool and on the means of ensuring a correct return of the information contained in the document when the document is tilted or positioned in a non-optimal manner with respect to the scanning tool.

The document EP0444083 relates to a method that resolves this problem a priori. More specifically, it relates to the use of an integrated scanner within a computer mouse, for obtaining data relating to the elements of an image on a page and mapping said data in the memory of a computer .

The method includes placing a scanner having a long row of elements and two optical mice covering an image on the page to be scanned, determining the orientation of the scanner with respect to the grid using the motion detection information by the two mice. , position the scanner at a starting point where mouse position detection begins, and start scanning the page by moving the scanner across the grid surface, and record the pixel data. digitized by each detector element to a memory address of a computer in communication with the scanner.

The method proposed in this document allows a retrieval of data presented by a document, regardless of the relative position of the scanner and said document and could for this reason satisfy the needs of an information record from a document that does not would not be positioned optimally with respect to a scanner. However, the step of careful scanning by the scanner of said document appears tedious implementation in the context of rapid contacts between people in a forum and its impractical application.

The problem of automatically saving specific data from a document in a simple, fast and modern way of implementation is not entirely solved by existing solutions.

Solution proposed by the invention

In order to remedy these drawbacks, the invention relates to a device for recording information from a capture of an image of a document, comprising: a terminal provided with an image capture objective, a screen and telecommunication means, - a support for holding the terminal on a support surface, in a predetermined position in which its screen points above the horizontal, - a reflection module mounted on the support provided with a reflective surface arranged opposite the objective of the terminal so as to form when the terminal is held by the support, an image of an area of the support surface located on one side of the support and viewable by the objective of capturing the image of the terminal, the terminal being provided with means for correcting the dimensions and perspective of the image of an object of regular shape disposed in said zone to best match it to the dimensions and the appearance of said object . The invention may furthermore exhibit one or both of the following features: the means for correcting the dimensions and perspective of the image of the zone obtained comprise a three-dimensional shape of the reflecting surface, the three-dimensional shape of the reflecting surface comprises a concave portion and a planar portion, the means for correcting the dimensions and perspective of the image of an object comprise computer tools for detecting the contour of the image of the object, identifying an edge of least deformation of the object and the opposite edge of the object, applying a corrective factor depending on the dimensions of the edge of least deformation and the opposite edge of the object applied to at least one part of the image of the object, - the computer tools include tools for identifying images and / or text present on the image of the object and for recording the images and / or text identified as that cont Specifically, the device may comprise computer objects that can be modified by a user and that can be associated with the specific contents recorded. The invention also relates to a system for recording information from a capture of an image of a document, comprising the device as described above and a remote server provided with a storage unit of the invention. corrected image obtained from the object communicated by the terminal.

According to another characteristic, the specific contents of the same object are recorded in the storage unit with the same identifier.

According to yet another characteristic, the specific contents of the same object stored in the storage unit are respectively associated with corresponding specific available locations of a file to be completed.

DESCRIPTION OF THE FIGURES The invention will now be described according to a possible embodiment, not limiting, with reference to the following figures: FIG. 1 represents a schematic side view of the device according to the invention for recording the data contained therein; on a sheet of A4 format, - Figure 2 which shows a front view of the device of Figure 1, - Figure 3 which shows a top view of the sheet whose data are to be recorded, according to an exemplary embodiment, FIG. 4 represents the image obtained by the terminal before the implementation of the size and perspective correction means of the device according to the invention, FIG. 5 represents the image corrected by the device according to the invention and annotated comments of a user of the device according to the invention.

DETAILED DESCRIPTION OF THE EMBODIMENT The invention relates to a device for recording information from a capture of an image of a document. As represented in FIGS. 1 and 2, the device according to the invention comprises a terminal, preferably mobile, of the tablet type 1 provided with an image capture objective 2, a screen 3 and telecommunication means (not shown) for remotely transmitting data to a remote server.

This tablet is held in a fixed position by a support 4 made of a rigid material allowing it to hold the tablet in a particular inclination and fixed at (60 ° in the example shown). According to the illustrated embodiment, this support has a V-shape in cross section, of which a shorter branch rests on the support surface 6 and makes an angle α with the second branch, which is longer and consists of a form frame. complementary terminal (rectangular in this case) in which the terminal is clipped. The dimensions of the base of the support 4 and the frame for the tablet allow it to maintain it in an inclined and stable position without other means of maintenance. From the front (Figure 2), the support 4 includes a frame surrounding a reception window and presentation of the screen of the tablet, also leaving the lens of its camera 2.

When the terminal is held fixed and locked next to the window of the support, it is inclined above the bearing surface of the angle a and points its screen (the normal of its screen) along an axis raised with respect to the horizontal, so as to be visible by a user whose eyes are arranged above the bearing surface 6.

The device according to the invention further comprises a reflection module 7 removably or permanently mounted on the support in order to reflect an image of a document 8 positioned in front of the support and from which it is desired to extract information. For this purpose, this module 7 is provided with a reflecting surface or mirror 9 arranged opposite the objective of the terminal so as to form when the terminal is held by the support in its fixed inclined position, an image of a zone of the support surface located on one side of the support and viewable by the image capture objective of the terminal.

This module is in the form of a U-shaped piece in cross section, which is mounted on the upper edge 11 of the support with reversible locking means on this edge 11 (for example magnetic and the terminal will be provided with a complementary magnetic element), a branch of the U internally carries the mirror 9 by positioning it with respect to the frame of the support. In the example shown, the reflecting surface is inclined relative to the inclined branch 10 of the support by an angle β less than the angle α, preferably less than 5 ° +/- 4 °, in this case angle β is 55 ° in the illustrated example.

When the terminal is mounted on the support and the module on the support, and a document 8 of type A4 sheet is disposed in front of the support (at any proximity thereof, for example with an edge touching said support) , a raw image is formed, without correction 12 of the document on the mirror 9, which captures the camera / camera of the terminal, with however a distortion due to the difference in distance existing between different points of the document with respect to of the objective 2, a function of the angle β and the height H of the objective with respect to the bearing surface 8. This height H is chosen to be greater than a threshold value, for example 20 cm. As shown in the schematic figure 4, the edge of the document farthest from the lens 12 appears smaller 5, the nearest one, larger 10 and the intermediate portions of the document becoming narrower as one away from the closest edge of the lens. The raw image of the rectangular document thus has a generally trapezoidal shape. Also, the resolution of the different parts of the image obtained is all the worse as one moves away from the edge of the image closest to the objective 2.

To overcome this problem, the invention provides that the terminal is provided with means for correcting the dimensions and perspective of the image to correct the dimensions and the resolution of the image obtained.

These correction means can integrate a particular three-dimensional shape of the reflecting surface 9 allowing it to enlarge an area of the document farthest from the lens and thus overcome the problem of sizing / resolution.

The reflective surface may be ainti having a flat shape for its portion reflecting the part closest to the document, and a concave portion for its portion reflecting the farthest part of the document.

According to another embodiment, this three-dimensional shape of the reflecting surface may have a continuously progressive concavity from a reflective surface end strip of zero concavity to a maximum concavity reflecting surface band, with a variation of concavity function of the deformation and / or loss of resolution obtained for a document of standard format (for example A4).

The search for the optimal progression of the concavity of the reflective surface can be carried out empirically to determine the radius of curvature and / or the width of a strip of the reflective surface to obtain the slightest deformation and / or loss resolution for the portion of the document reflected by this tape.

It is possible to provide a set of modules whose reflective surfaces would be of different three-dimensional shapes in order to correct the deformations and / or loss of resolution occurring during the capture of an image of documents of different standard formats (A4, A3, business cards). ...) for the same medium.

The correction of the dimensions and / or the resolution can be further performed by computer tools.

These computer tools may include tools for detecting the outline of the image of the object, identifying an edge of least deformation of the object and the opposite edge of the object, application of a factor corrective function of the dimensions of the edge of least deformation and the opposite edge of the object and applied to at least a portion of the image of the object, for example the one closest to the edge of least deformation of the object.

This corrective factor may of course be progressive between the two opposite edges of the document to obtain a better correction.

To exploit directly when shooting the document, its content, the invention provides tools for identifying graphics and / or alphanumeric characters present on the corrected image of the object and recording images and / or text identified as specific content.

In order for a user to associate comments with a captured document, the invention provides a user interface for modifying computer objects and associating them with the specific contents recorded for a given document and thus for a given candidate.

Specific content (photo, e-mail address, mailing address, age, experience, training ...) and the recruiter's comments (additional information, impressions, first opinion) can be transmitted by the terminal to a remote server provided with a unit for storing the corrected image obtained from the document and specific contents and associated comments, under the same identifier, in the form of a database. The user, recruiter, may be provided with his own terminal with an interface dedicated to the formatting of information (for example taking the form of a form to be completed by a candidate wishing to apply for a particular advertisement), which can automatically orderly extract the data of each candidate registered in the database corresponding to this announcement and automatically fill in as many application forms as candidates who have applied for this announcement.

The device according to the invention may if necessary, be coupled to a personalized intelligence intelligence or "chatbot" replacing the recruiter on a forum or the application procedure for an announcement, chatbot who will interview the candidate in the form of questions in a proposed interface on the terminal after capturing and identifying the contents of the candidate's document, in order to direct the questions of the chatbot according to the information collected. This interview with virtual recruiter can allow as described above, to automatically fill out application forms in the interface of the recruiter's terminal.

According to a variant not illustrated, the module may be provided with a steerable reflective surface or the branch of the module supporting this reflecting surface may be orientable, preferably automatically. Also, the inclined portion of the support may be provided automatically orientable.

It will then be possible to provide an initialization phase of the module and / or the support, consisting of seeking the optimal orientation of the reflecting surface and / or the inclined part of the support enabling it to form an uncorrected image of a document of calibration, from which the image corrected by the device according to the invention will be as faithful as possible of the calibration document.

The module can be made intelligent by integrating a microprocessor equipped with computer correction tools, a storage unit of the collected information and telecommunication means with the terminal and / or the remote server.

Claims (10)

claims A device for recording information from a capture of an image of a document, comprising: a terminal (1) provided with an image capture objective (2), a screen (3) and telecommunication means, - a holding support (4) of the terminal on a bearing surface (6), in a predetermined position in which its screen points above the horizontal, - a module of reflection (7) mounted on the support (4) provided with a reflecting surface (9) arranged facing the objective (2) of the terminal so as to form when the terminal is held by the support, an image (12) an area of the support surface situated on one side of the support and viewable by the image-capture objective of the terminal, the terminal being provided with means for correcting the dimensions and perspective of the image of the image an object of regular shape disposed in said area to best match the dimensions and appearance of said object. 2. Device according to claim 1, wherein the means for correcting the dimensions and perspective of the image of the area obtained comprise a three-dimensional shape of the reflecting surface (9). 3. Device according to claim 2, wherein the three-dimensional shape of the reflecting surface (9) comprises a concave portion and a flat portion. 4. Device according to claim 2, wherein the three-dimensional shape of the reflecting surface (9) has a concavity of progressive radius of curvature. 5. Device according to any one of the preceding claims, wherein the means for correcting the dimensions and perspective of the image of an object comprise computer tools for detecting the outline of the image of the object, identifying an edge of least deformation of the object and the opposite edge of the object, applying a corrective factor depending on the dimensions of the edge of least deformation and the opposite edge of the object applied to at least one part of the image of the object. 6. Device according to the preceding claim, in which the computer tools comprise tools for identifying graphic and / or alphanumeric contents present on the image of the object and for recording the graphic and / or alphanumeric contents identified as specific content. 7. Device according to one of the preceding claims, comprising computer objects that can be modified by a user and associated with the specific contents recorded. 8. A system for recording information from a capture of an image of a document, comprising the device according to one of the preceding claims and a remote server provided with a corrected image storage unit. obtained from the object communicated by the terminal. 9. System according to the preceding claim, comprising the device of claim 6 or 7, wherein the specific contents of the same object are recorded in the storage unit with the same identifier. 10. System according to the preceding claim, wherein the specific contents of the same object stored in the storage unit are respectively associated with corresponding specific available locations of a file to be completed.