DE102008060034B4 - Storage device with documentation of the removal of objects and corresponding storage and removal procedure - Google Patents

Abstract

A storage device for storing articles with a frame structure (1) and at least one storage unit (2) movable relative to the frame structure in two opposite directions (forward and return directions), having a storage surface (L) for supporting at least one article thereon, an image pickup unit (10, 11) with which the storage surface of the at least one storage unit at its back and during their return movement relative to the frame structure is at least partially detectable and a connected to the image pickup unit and / or integrally formed with this image evaluation unit (16) by the Image recording unit detected image data are transmitted and which is adapted to generate a Hin-documentation image from at least one time during the forward movement by the image acquisition unit detected Hinbewegungs image data and for generating a re-documentation image from at least one em point during the return movement detected by the image pickup unit return motion image data.

Description

The present invention relates to a storage device for storing objects, such as utensils such as tools or even consumables such as food, with the removal of objects from the storage device is documented. The invention also relates to a corresponding storage and / or removal method and to a documentation method for these operations.

In order for a smooth workflow to be achieved in business operations, consumables and consumables needed must be held up according to the (expected) wear and / or consumption. In manufacturing companies, for example, this applies to interchangeable tools (drills, grinding wheels, etc.), and in hospitals to medication packages.

From the point of view of the organization of work, such stocking systems are particularly suitable for the provision, which allow an independent removal by a requester, since in this way cost-effective high availability of the goods can be realized.

For self-contained storage systems, the following two conditions should be met, depending on the application scenario: On the one hand, this is personal access control, if the issue of the goods may only be performed by authorized persons (eg delivery of medicines) or if the costs are directly charged must be assigned to extracting persons (for example, when billing removed tools broken down by organizational units). On the other hand, a continuous documentation of the inventory may be required in order to be able to automatically trigger repeat orders on this database.

With so-called automatic dispensers, both the authentication of the requester and the documentation of the stock can be implemented technically well. In the case of a few objects to be kept, machines with drawers (example: cigarette machine), machines with spiral compartments (example: sweets machine) or machines with doors (example: bottle vending machine) are used. These machines have the advantage of allowing complete inventory control; the requestor only approaches the desired items via a mechanical release. An authentication of the requester can also be easily realized via an access code, a magnetic card or a key. The disadvantage of these systems, however, is a high space consumption, since per base unit only a few different items are vorhaltbar.

Another device for holding objects are so-called drawer cabinets. Drawer cabinets consist of several superimposed drawers, which in turn are usually divided into different compartments. Drawer cabinets have the advantage that many different items can be stored with low space consumption. The individual compartments can be equipped individually depending on the application scenario, with one or more copies of an object (for example, a drill type with a specific size and special purpose). An access control can be realized via devices that allow locking of individual drawers or the entire cabinet. A release can then be implemented by the already mentioned authentication measures.

However, drawer cabinets are problematic in terms of documentation of the removal processes and stock levels. This is because the requestor not only has access to the compartment containing the desired item, but also to all previous compartments. Therefore, it can not be ruled out that the requester removes too many items or even wrong items, be it by mistake or intentionally. This leads to discrepancies between the real stock (defines the actual state) and the associated stored data (defines the target state).

If discrepancies occur during inventory control, for example as part of a refilling process, it is not possible to clarify this fact at first. In particular, it is not possible to trace back any discrepancies to concrete erroneous removal processes (with date, time and, in the case of an authentication, with the identity of the requester).

• • Mit Hilfe einer steuerbaren Verriegelung lassen sich Schubladen nur soweit ausziehen, bis das Abteil mit dem gewünschten Gegenstand erreichbar ist (Patentschrift DE 699 06 299 T2 ). Damit sind die dahinter liegenden Abteile einer Schublade vor einer Manipulation geschützt; nicht jedoch die davor liegenden.
• • Mit Hilfe eines Spindel- oder Zahnstangenantriebs wird eine Schublade nur soweit ausgefahren, bis das Abteil mit dem gewünschten Gegenstand frei zugänglich ist (Patentschrift EP 1 813 168 A1 ). Damit sind ebenfalls nur die dahinter liegenden Abteile vor einer Manipulation geschützt; nicht jedoch die davor liegenden.
• • Bei einer anderen Herangehensweise werden die Abteile in einem Schubladenschrank jeweils einzeln durch oben aufliegende Klappen geschützt (kommerzieller Anbieter: Firma W. Achterberg in Hamminkeln-Brünen). Über eine zusätzliche Vorrichtung zur Verriegelung können diese Klappen einzeln gesperrt bzw. entsperrt werden. Ebenso ist es möglich, über einen zusätzlichen Sensor pro Abteil das Öffnen und Schließen der Klappen zu dokumentieren. Bei diesen Systemen ist eine vollständige Dokumentation möglich, sofern sich in den Abteilen jeweils nur ein Gegenstand befindet. Sind hingegen mehrerer Gegenstände in einem Abteil enthalten, so kann es weiterhin zu Fehlern bei der Entnahme kommen. Die Installation der Klappen mitsamt Aktorik und/oder Sensorik ist zudem apparativ sehr aufwendig.

In order to allow an improved control of the stock, several solutions are provided in the prior art:

• • With the aid of a controllable lock, drawers can only be extended until the compartment with the desired object can be reached (Patentschrift DE 699 06 299 T2 ). Thus, the underlying compartments of a drawer are protected from manipulation; but not the one in front of it.
• • With the help of a spindle or rack drive, a drawer is only extended until the compartment with the desired object is freely accessible (Patentschrift EP 1 813 168 A1 ). This is also only the behind compartments protected against manipulation; but not the one in front of it.
• • In another approach, the compartments in a drawer cabinet are individually protected by overhead flaps (commercial provider: W. Achterberg in Hamminkeln-Brünen). About an additional device for locking these flaps can be individually locked or unlocked. It is also possible to document the opening and closing of the flaps via an additional sensor per compartment. Full documentation is possible with these systems, as long as there is only one item in each compartment. On the other hand, if several items are contained in one compartment, it may continue to come with errors in the removal. The installation of the flaps together with actuators and / or sensors is also very expensive in terms of apparatus.

From the DE 297 24 869 U1 is a storage device with a storage room and with a closable access opening to the storage room known. The storage device is associated with an optical imaging device serving to detect a three-dimensional object, the generated images of which are processed by screen technology and fed via data lines to an electronic visualization device and to the data processing of a data processing device.

In addition, the describes DE 32 08 135 A1 a method for the detection and identification and / or quality control of objects, in particular of books or the like., Which are preferably subjected to storage and retrieval of a length, width and height measurement and provided with a readable bar code. By adding up the brightness values line by line and / or column by column, a characteristic linear brightness range of the surface of an object can be generated.

Based on the described prior art, it is thus the object of the present invention to provide a storage device for storing objects, in particular a drawer cabinet system, with which inexpensive; simple and reliable way a complete documentation of the removal of objects from storage units, especially drawers, is possible. It is also the task to provide a suitable storage and / or unloading procedure as well as a documentation procedure for these processes.

This object is achieved by the storage device according to claim 1 and by the storage and / or removal method according to claim 24. Advantageous embodiments of the storage device according to the invention and of the storage and / or removal method according to the invention can be found respectively in the dependent claims.

The storage device according to the invention as well as the storage and / or removal method according to the invention and the documentation method will first be described generally below. This is followed by a detailed special embodiment. However, the configurations specifically described in the exemplary embodiment and in the general description need not be realized in this form within the scope of the invention. The features of the present invention, as defined by the claims, may be embodied within the skill of the artisan also in other combinations of features.

The basic idea of the present invention is to image the storage area of the individual storage units (usually: drawers) of the storage device with an image acquisition unit. As a rule, a plurality of storage units or drawers are arranged one above the other in a stack arrangement. The image pickup unit is preferably mounted inside the bearing device, so that the image pickup can not be influenced from the outside. For a storage area in use (i.e., extracted or moved out by a user), images are generated when at least partially visible to the imaging unit. By a mechanical or electrical safety device is thereby usually ensures that only ever exactly one storage unit or drawer is in use. According to the invention, images (in the following: documentation images) which document the state of a storage surface (that is, for example, the position of the objects in a drawer) are composed of image sections acquired at different times or in different states (eg drawer positions).

In the case of a forward movement (eg during a pull-out movement of a drawer), the corresponding storage area (eg interior area of a drawer) is detected section by section with the camera and the detected sections are suitably assembled, according to the invention a first documentation image (hereinafter: first Documentation image or out-of-documentation image). This back-documentation image then documents the state of the storage surface (that is, for example, the interior of a drawer with the objects stored therein) prior to removal of an object by a user. Likewise, in a second, essentially opposite directed movement (return movement), that is z. For example, when reinserting a drawer pulled into a drawer cabinet, also the storage area in sections, ie at different times or in different positions of the currently scanned storage unit relative to the surrounding, supporting structure (frame structure, eg., Housing the drawer cabinet with rails to movable fixing of the individual drawers) scanned. From these image data acquired in sections during the return movement, a further documentation image (hereinafter: second documentation image or also back-documentation image) is then produced by assembling similar to the forward movement. This back-documentation image then documents the state of the scanned storage area after a user, e.g. B. has taken an item from the corresponding storage area. By suitable comparison of the two documentation images, that is, the outward and the back documentation image, it is then possible to identify removed objects. The latter can be realized, for example, computer-aided by suitable image processing algorithms (eg by the blob analysis described below in more detail or also by pattern matching algorithms).

It is thus essential that a storage device according to the invention initially has a frame structure and at least one (generally several) relative to this frame structure in two substantially opposite directions (Hinbewegungs- and Rückbewegungsrichtung) movable storage unit each having a storage surface for storage of objects thereon. The frame structure may, as already mentioned, be the housing of a drawer cabinet, with the storage units arranged around the individually arranged drawers with their inner surfaces as storage surfaces.

The term "substantially oppositely directed movement" (or the direction of return and the direction of return) is understood in the context of the present invention in a very general way: it does not have to be (spatially) generally a mere translation movement (such as It usually occurs in drawer cabinets act), but the movement can also include rotation components. It is essential then that the forward movement direction has at least individual sections which are directed opposite to individual sections of the return movement direction. Also, the terms of Hinbewegungsrichtung and the return direction are seen in terms of time very general: So the forward movement corresponds to such a drawer system not only just a pure pull-out movement of a single drawer, d. H. a movement in which the drawer is pulled out only ever further from the corresponding frame structure during a time interval, but also includes movements that are designed during the time interval in which the corresponding drawer is moved, only predominantly as pull-out movements. In other words, according to the invention, a movement in which a user pulls out a drawer around a first section from the associated drawer cabinet, the drawer then pushes back a small piece in the cabinet, then finally as far as possible to pull it out of the drawer cabinet also understood as a forward movement. The same applies to a return movement (predominant insertion of the drawer), in which the main direction of movement (return movement) so may well have the actual return motion interrupting Hinbewegungsanteile.

The storage device according to the invention has an image recording unit, in particular a camera, with which the storage area of the individual storage units during the forward movement and during their return movement (relative to the frame structure) can be detected in a pictorial manner. In this case, the pictorial detection for the reasons described above is usually done in sections and usually by detecting strip-shaped bearing surface sections perpendicular to the direction of return and return movement. As described above, the image recording unit is preferably arranged within the frame structure (drawer cabinet housing).

An image evaluation unit (eg in the form of a PC with corresponding software), to which the image data or camera data acquired by the image acquisition unit can be transmitted, is connected to the image acquisition unit. The image evaluation unit then assembles the first documentation image (out-of-documentation image) from the different sections of a storage area recorded at different times during a forward movement. Likewise, it uses the different surface sections of the storage area, which are recorded at different times during the return movement, to compile the reverse documentation image. Each documentation image is in this case composed of a plurality of sections of the storage area, which are depicted in pictorial form at different times, exactly one storage unit.

In a particularly advantageous variant of this composition takes place in the two documentation images so that only for the first time in the corresponding movement occurring image sections in the documentation image composed become: Thus z. B. a first movement to pull out a single drawer to half of the maximum pull-out, which is then interrupted by a short, oppositely directed insertion until the withdrawal is finally continued until the end (drawer fully open), not all of The camera used during this movement frames used to assemble the Hin-documentation image: Until the end of the first part of the extraction movement all captured surface portions occur for the first time, are thus used for the Hin documentation image. During the interruptive return movement, individual surface sections are scanned again. These resampled area sections are then not used for the documentation image. Only when further surface sections occur for the first time during the final, renewed pull-out movement, are these incorporated again into the background documentation image.

The control of which surface sections have already been detected and which are new, as described in more detail below, takes place with the aid of position elements (for example strips attached to the drawers) which have a large number of position markings (for example binary coded position markings).

In a particularly preferred variant, a part of the device according to the invention, namely the device part comprising the image recording unit and image evaluation unit, can also be designed as a retrofit kit: This device part is then designed so that it can be placed on an already existing drawer cabinet, so that the present invention can also be used within existing storage facilities (retrofitting of existing systems).

Advantageously, a storage device according to the invention can be provided with an authentication system (which, for example, can also be embodied in the above-described PC). It is then a personal access control connectable with a documentation according to the invention: If it is detected which persons access the storage device at which time points, removed objects identified by means of the documentation images can be clearly assigned to the corresponding users.

Likewise, the present invention may be provided with a documentation system for the documentation of inventories by identifying objects removed (or also added by evaluation of documentation images created at suitable times) on the basis of the evaluation of the two documentation images.

The present invention will now be described in detail with reference to an embodiment.

Showing:

1 a cross section through a drawer cabinet according to the invention in the direction of the stacking direction of the individual drawers.

2 a sectional view through a drawer cabinet according to the invention parallel to the levels of the individual drawers.

3 an inventive position element, as it is arranged on each drawer of the drawer cabinet according to the invention.

4 a variant of in 1 shown drawer cabinet, in which this was formed by means of a retrofit kit.

5 the structure of a back and forth documentation image according to the invention.

6 a sketch for the identification of stored in a drawer items by means of suitable image processing algorithms

1 shows a cross section in the direction of stacking by a storage device according to the invention in the form of a drawer cabinet. The drawer cabinet has a frame structure 1 in the form of a drawer (in the inserted state) surrounding housing. Inside the frame structure 1 are parallel to each other but offset in height, a plurality of rails (not shown) on opposite walls of the frame structure 1 educated. There are four individual drawers on these rails 2 movably fixed. The individual drawers are arranged one above the other in the form of a stack. Each individual drawer can be almost completely removed from the frame structure in the direction parallel to its fastening rail 1 of the drawer cabinet. In this case, a mechanical or electrical safety device (not shown) is provided so that only exactly one of the drawers of the frame structure 1 can be pulled out.

As 2 in a cross section perpendicular to in 1 shown cross-sectional area are the drawer frame 3 the individual drawers designed so that (in the shown, fully inserted state of the drawers) which enclosed by them, designed for the storage of objects storage surfaces L of the individual drawers (from the center of the Drawer cabinet seen from) not to the front F (front of the drawers 2 can be pulled out) of the frame 1 are extended, but already at a distance D from this front end. This distance D becomes a gap 6 between the front F and front wall of the frame structure 1 and formed between the front surface of each drawer. This (air-filled) gap 6 is also referred to below as a free field of view 6 designated. As a connection between the frame 3 a drawer and the outward, by a user, the drawer from the frame structure 1 wants to pull out, tangible front panel 4 the drawer and handle are two with the front-front corners of the drawer frame 3 connected bars 5 intended.

Like now 1 is above the top drawer within the frame structure 1 an image pickup unit in the form of a CCD or CMOS line scan camera 10 with a suitably trained, upstream lens 11 arranged. The camera is here above the rear end of the top drawer inside at the top O of the frame structure 1 arranged and has, since it is located above the upper edge of the uppermost drawer and is aligned parallel to this edge, free field of view up to the front end F of the frame structure 1 ,

Above the free field of view 6 and above the footbridges 5 is a line-shaped mirror 12 arranged, which extends over the entire inner width B of the frame structure 1 extends (cf. 2 ). The mirror 12 is tilted to the horizontal or to the upper edge of the drawer frame 3 arranged so that the total height H of the free field of view (in 1 for reasons of clarity laterally offset outside the frame structure drawn) on the lens 11 sharp on the camera chip of the camera 10 can be mapped.

As 2 shows, has each of the four drawers on one longitudinal side of a drawer frame 3 fixed, horizontal position element 7 in the form of a bar. The bar 7 thus runs parallel to the mounting rail of the respective drawer and extends almost over the entire inner region of the frame structure 1 from its front F to its opposite rear R. On the top (the camera 10 facing side) carries the position element 7 a variety of position markings 8th (compare 3 ) in the form of a binary pattern. This binary pattern 8th is on the position element 7 arranged so that when successively pulling out a drawer each different sections of the binary pattern through the free field of view 6 or be detected by the camera detectable gap. As described in more detail below, depending on the position of the drawer is so relative to the frame 1 a different section of the position element 7 (and thus a different position marker) detected, so that by appropriate image analysis, the current position of a drawer 2 relative to the frame structure 1 , that is, their pull-out state, can be determined.

As 1 further shows is inside the frame structure 1 at the rear side R a with the image pickup unit or camera 10 connected image evaluation unit 16 in the form of a PC with suitable software. The from the camera 10 captured image data thus become the image evaluation unit 16 which, as will be described in detail below, generates a backward documentation image from a forward movement and a backward documentation image from a backward movement.

The computer 16 is provided with a storage and evaluation unit (not shown) for storing the back and forth documentation image and for evaluating the same by means of digital image processing algorithms described in more detail below. By means of these image processing algorithms, in particular the outward and the rearward documentation image can be evaluated with respect to objects removed from the back-documentation image but no longer imaged by the back-documentation image, ie removed from a corresponding drawer during and / or after removal.

As will also be described in more detail below, empty images (ie images of the individual interior drawer inner surfaces L without objects stored thereon) can first be generated for evaluation: If a drawer has, for example, a compartment or fan structure, the corresponding compartment walls can be eliminated by means of the empty images so that there is no confusion with stored items. Such storage surface structures (compartment and / or fan structures) of the storage areas of the individual drawers can also be stored and kept in advance in the storage and evaluation unit. The distinction between storage surface structures and stored objects can be made in the simplest case by means of simple formation of difference images.

The starting point of this embodiment is thus a drawer cabinet with several drawers 2 in which, due to mechanical or electrical precautions, only one drawer can be pulled out at a time, and whose essential components are constructed as follows:

Construction of drawers:

2 shows the view from above on one of the drawers 2 in the drawer cabinet or its frame structure 1 are installed. The drawer frame 3 are not guided to the outside with closed drawers, but via two each externally mounted webs 5 each with a front panel 4 connected. Each front panel is here provided with a handle. In this way, the free field of view results 6 between the front panel, the drawer frame and the bars when looking at the drawer as shown in this illustration.

Furthermore, the side of the drawer frame is a bar 7 with positioning marks 8th appropriate. The positioning marks are such that with them the opening state of the drawer can be determined, as will be explained later in the description of the method in more detail. In a typical embodiment of the invention, a black and white binary pattern is applied to the bar as a marker, as shown in FIG 2 and 3 is apparent. The binary pattern encodes in the columns numbers from 0 to n in the usual binary coding, in which 2 and 3 each ascending from left to right. The number of rows is determined directly from the largest number n to be coded. For the coding of n = 127, for example, 7 bits are required, the binary pattern therefore has 7 rows (see FIG 3 ). The number coded in a column of the binary pattern is called binary code in the following.

In addition to the binary pattern is on the front, ie the front side F facing the end of the position element a recess 9 in the position element 7 intended. This recess 9 used to identify whether at least one drawer has been pulled out. Once this is the case, the process of the documentation program begins.

3 shows a bar 7 with position markings or position marks 8th in the form of 7-bit coded numbers from 0 to 127.

Optionally, the binary pattern can also be distributed on two bars, with the second bar attached to the opposite side of the drawer frame (compared to the first bar).

Devices in the drawer cabinet or within the frame structure 1 :

The drawer cabinet has a camera above the drawers 10 with lens 11 attached (see 1 ). In this typical embodiment of the invention, the camera is a line scan camera with a line width of 2048 pixels and a refresh rate of 1 kHz. Furthermore, the mirror is located directly above the free field of view 12 which extends over the entire width of the cabinet.

With the chosen geometric arrangement of the components 10 to 12 the camera captures the free field of view 6 between the drawers. The camera lens is designed so that the entire area between the mirror and the bottom of the drawer cabinet is sharply imaged on the camera chip.

With the selected geometric arrangement is further ensured that in a single camera image, the strips 7 for all drawers are visible at the same time, as long as all drawers are closed (resulting from the opening angle of the lens and the associated perspective). In this case, the camera will see the binary code 0 on all bars.

If a drawer is pulled out of the frame structure, the associated binary codes 1, 2, 3, etc. (= columns of the bit pattern) are successively imaged on the camera chip. From the location of the binary codes in the camera image can be deduced based on the perspective of the recording on which drawer was pulled out. As soon as a drawer is pulled out, all underlying drawers (including the binary codes) are concealed.

Above the free field of view 6 is a switchable lighting 13 appropriate. This serves to illuminate the image area that is captured by the camera; in particular, it illuminates the strips with the positioning marks.

Here is (optional) also a light sensor 15 on the bottom plate U of the drawer cabinet and another lighting 14 on the cover plate O of the drawer cabinet integrated in the structure (or vice versa). Both elements are positioned so that when the cabinet is closed (all drawers in the cabinet) there is a direct line of sight between the light sensor and the additional lighting. The direct line of sight is thereby over the recesses 9 all strips installed in the drawer cabinet. As soon as a drawer is opened, the direct line of sight is interrupted.

In the drawer cabinet, the image evaluation unit is in the form of a computing unit 16 installed (called calculator in the following), with which the images of the camera are read out and the calculation of the documentation images is made. The computer controls the lighting and, if necessary, reads out the status of the additional light sensor. The computer can optionally manage the stock of the drawer cabinet (documentation system). The computer can optionally carry out the evaluation of the documentation images. The computer may optionally have an input and output unit (authentication system) through which the requester can authenticate, via which the requestor's request for an item is received and over which the results of the image analysis are displayed.

As 4 shows, the present invention can also be implemented in the context of a retrofit system, in which a separate, the image pickup unit 10 . 11 and the image evaluation unit 16 comprehensive device part is designed as a retrofittable element (see the cross-sectionally L-shaped, attached to the top of an existing drawer system device part 17 in the 4 ).

The basic structure of using the retrofit element 17 trained drawer system of 4 is as described above, with the following differences: In the present case, to ensure the same height of the drawer cabinet as in the 1 shown, the existing drawer element part with only three individual drawers 2 fitted. This part comes complete with its drawers 2 simply adopted in the manufacturing form, as the only change must be to the frame of each drawer 3 each bar 7 with position markings 8th be attached. All other additional components described above are in the second, designed as an attachable retrofit kit housing 17 housed, which is directly attached to the actual drawer cabinet 1 connects and is firmly connected to this.

The two documentation images are in the in the 1 to 4 shown variants of the storage device according to the invention generated as follows:
Each time a drawer is used by a user on a drawer, two documentation images are produced, one outward image image during a forward movement (pulling out a drawer) and one backward documentation image during the return movement (pushing in the drawer). The first documentation image (hereinafter also referred to as image n1) documents the state of the drawer at the beginning of the extraction process, the second documentation image (hereinafter also called image n2) documents the state of the drawer at the end of the process (ie when the drawer is pushed back into the frame structure has been). The image evaluation unit draws this 16 first camera images of the camera 10 during opening and closing of the relevant drawer 2 The composition of the two documentation images is then carried out on the basis of the quantity of recorded camera images. This will now be described in more detail below.

At the beginning of an image capture by means of the camera 10 are all drawers 2 closed. In a first variant (without additional lighting 14 and light sensor 15 ) is the lighting 13 switched on in the drawer cabinet. The camera 10 continuously records individual camera images with a defined image acquisition frequency and sends them to the image evaluation unit 16 , Once when opening a drawer 2 in the camera picture a binary code (compare 3 ) greater than 0 (by pulling out the corresponding bar 7 ), the following sequence of image recordings of the camera is stored in a memory of the image evaluation unit or the computer 16 saved. Image capture by the camera stops when the binary pattern on the bar can no longer be detected by the camera (ie when the drawer is fully retracted).

In another variant (with additional lighting 14 and light sensor 15 ) will open the drawer via the additional light sensor 15 detected, since in this case the direct line of sight between the lighting 14 and the light sensor 15 is interrupted. The lighting 13 is then turned on and the following image sequence of camera images in the computer 16 saved. Image capture by the camera stops when the additional light sensor 15 Detected the closure of the drawer. Subsequently, the lighting 13 switched off. At the end of the picture taking by the camera are all drawers 2 closed.

5 sketched as a documentation image (Hin-documentation image or back-documentation image) is constructed: The figure shows a pixel grid of a digital, in the memory of the PC 16 storable image, which here z. B. has a size of 2048 × 2048 image pixels (ie PZ or the number of pixels per image column is 2048 and thus corresponds to the number of CCD or CMOS cells of the line scan camera 10 ).

In the following, the generation of an out-of-documentation image is described, a back-documentation image arises accordingly. When pulling out the drawer captures when the first on the bar 7 arranged binary code 8 (hereinafter called BC1) by mapping the gap 6 over the mirror 12 and the lens 11 on the camera chip of the camera 10 is imaged, the camera a first strip-shaped portion of the storage area L of the drawer pulled out 2 , Is the picture-taking frequency of the camera sufficiently high, so while still pulling out the first binary code BC1 of the bar 7 is detected, a plurality of individual camera images K of closely spaced strip-shaped sections on the front surface F facing side of the storage area L of the drawer 2 detected.

In the in 5 In the case illustrated by way of example, these are four individual camera images K1 to K4, which are detected during the withdrawal of the binary code BC1. Since four image columns per binary code are available in the documentation case in the illustrated case, the four individual camera images K1 to K4 of the first binary code BC1 are written into the first four columns of the rear documentation image.

The in 5 The case illustrated represents a simple example for the sorting of the camera images in the documentation images. The general procedure for the sorting of the captured camera images (in particular how they are averaged and / or summarized if necessary, adapted to a predefined width or number of columns of a documentation image to be and how, for example, with occurring during the pulling and pulling movement of the drawer, in the opposite direction directed return movement sections is handled) will be described below.

After taking the fourth camera image K4, the drawer is pulled out so far that the first time the second binary code BC2 on the bar 7 is detected. Then, assuming a constant pull-out speed, during detection of the second binary code BC2, four individual camera images, the images K5 to K8, are detected again, which in turn are written column by column into the columns 5 to 8 of the out-of-documentation image. The process continues until the out-of-paper image is completely filled.

Preparation of the documentation images:

The size of the documentation images is freely adjustable. It is typically specified when setting up the drawer cabinet. In a typical embodiment of the invention, a documentation image has as many pixels in height as the line scan camera used; Each pixel of the line scan camera is then directly associated with a line of documentation images. In a typical embodiment of the invention, the number of pixels in width is selected to give the documentation images an aspect ratio that corresponds to the natural aspect ratio of the drawers (length to width). The number of pixels in the width will then usually be greater than the number of numbers realized in the binary pattern, which is why each binary code is assigned several columns in the documentation images (for example, binary code 1, columns 1 to 4, binary code 2, columns 5 to 8, etc.).

The number of pixels in width is determined when the drawer cabinet is put into operation. The drawers all have the same width in the documentation screens because there is a direct mapping of binary codes to the column numbers in the documentation screens. However, the height of the drawers in the documentation images is variable: due to the perspective, the upper drawers have a high height and the drawers at the bottom have a low height. Accordingly, the aspect ratio changes a little (the aspect ratio of the drawers in the documentation images, while the aspect ratio of the documentation images is always the same). In order to achieve a true-to-reality image of the actual conditions, the width of the documentation images is set in the example described in such a way that an aspect ratio for a middle drawer is obtained which corresponds to the aspect ratio of the real drawer; it is slightly larger for higher drawer sizes. a bit lower for lower drawers.

Analysis of the camera images and calculation of the associated image columns:

The computer 16 analyzes the recorded camera images successively from the first to the last shot and assigns you a column in the documentation images according to the detected binary code.

In a typical embodiment of the invention, the assignment of the image columns takes place as follows: First, the recorded camera images are subdivided into temporally successive sequences. A sequence change always takes place when two adjacent camera images in the image sequence have different binary codes. The sequences thus have the property that the camera images contained therein all have the same binary code. The images of a sequence are assigned column information if and only if the binary codes of the previous and subsequent sequences differ (example of different binary codes: the images of the sequence under consideration have the binary code 4, the sequence before has the binary code 3, the sequence thereafter the binary code 5). In this way, the assignment of a column information is omitted when a camera image was recorded immediately before or after a change of the direction of movement of the drawer (change from an opening to a closing movement and vice versa).

The assignment of the camera images of a sequence to the image columns of the documentation images takes place according to the following scheme: The camera images of the sequence are distributed linearly uniformly on the available image columns (ie, on the image columns to the associated binary code). If, for example, a total of k camera images have been recorded for a single binary code, each individual camera image i is first assigned the value bf from the set {1..k} in a typical embodiment of the patent: bf = b1 + (b2-b1) · ( i - 1) / (k - 1). In this case, b1 designates the smallest or b2 the largest image column number for the specific binary code (a specific binary code is typically assigned a plurality of image columns, but at least one image column). The number of the image column for each camera image is given by (commercial) rounding of the number bf to the next whole number. For highly accurate image generation, optional decimal places can also be taken into account when calculating the column numbers.

The actual generation of the documentation images (or the filling of the individual image columns of in 5 Now, only the first occurrence of a defined position on the storage surface L may be taken into account. For example, when creating the out-of-picture image n1, a strip-shaped portion of the storage area L detected by the camera is considered for the out-of-picture image only when the associated binary code BC has occurred for the first time. If, on the other hand, it occurs again (after a binary code change), then the corresponding camera image (s) is no longer taken into account. For example, if the user pulls on the drawer, the second binary code BC2 first appears after a certain pull-out path , the corresponding camera images are taken into account. The user then briefly interrupts the pull-out movement by re-inserting the drawer into the frame structure 1 (ie the first binary code BC1 reappears) and by continuing the pull-out movement again, the binary code BC2 now appears for the second time, ie it is no longer taken into account. Only when still further pulling out (when the third binary code BC3 occurs) are the corresponding camera images taken into account again for the background documentation image.

For the documentation image n1 (documentation image that documents the state before a removal process), the individual recorded camera images are thus successively processed from the first to the last camera recording. At the beginning of the removal process, the drawer is completely in the frame structure 1 retracted, so no binary code is visible. If a certain binary code BC occurs for the first time (ie if a binary code is detected which is larger than the corresponding binary code values of all previously acquired camera images), the associated camera images are entered into the associated image columns in 5 accepted. Depending on the ratio of the frequency of the camera image acquisition and the speed of the extraction movement, it may happen that a column of the pixel matrix in 5 several camera images are assigned (the individual images are then averaged, for example, and the mean value is written to the corresponding image column). If image columns are not used when creating a documentation image, they can optionally be interpolated from adjacent columns. This gives a picture of the initial state of the drawer.

For the corresponding back-documentation image n2 (documentation image which describes the state after a removal process), the recorded camera images are then processed in reverse order from the last to the first acquisition and processed accordingly. At the end of the removal process, the drawer is completely in the frame structure 1 retracted, so no binary code is visible. If a specific binary code BC occurs for the first time during the successive passing through of the images from the last recording to the first recording (ie if a binary code is detected which is larger than the corresponding binary code values of all previously acquired camera images), the associated camera images are transferred to the associated image columns in FIG 5 accepted. Accordingly, this results in an image of the final state of the drawer.

Optionally, even decimal places of the calculated image columns can be taken into account when copying into the documentation images by first using a higher image resolution for the image generation, but finally converting the image to the set target image size.

The two documentation images generated for each opening process are permanently stored by the computer on a storage medium. The date and time are saved for each documentation screen. The documentation images can be analyzed immediately or at a later date.

Optionally, the requested object or also the name of the requesting person can be stored for each documentation image if the computer has information about this (information on the request or for authentication).

The documentation screens can be evaluated as follows:
The removal operations can be completely reproduced based on the calculated documentation screens. By evaluating the images, it is possible, in particular, to identify faulty removal processes (in short: erroneous withdrawals).

• • (H1) Erzeugung von Leeraufnahmen: Es wird für jede Schublade eine sog. Leeraufnahme erzeugt, die dadurch definiert ist, dass sich bei der Bildaufnahme kein Gegenstand in der Schublade befindet. Zusätzlich werden dabei die Skalierungsfaktoren der Bildaufnahmen bestimmt (in mm pro Pixel für jede Schublade). Die Aufnahmen werden in einer typischen Ausführung der Erfindung bei der Einrichtung des Schubladenschrankes erzeugt. Mit der Hilfsfunktion (H1) werden die Leeraufnahmen und die Skalierungsfaktoren den nachfolgenden Auswertungen bereitgestellt.
• • (H2) Hinterlegung der Abteilstruktur (Lagerungsflächenstruktur) für jede Schublade: Mit der Hilfsfunktion (H2) wird die Aufteilung jeder Schublade in voneinander abgetrennte Abteile bereitgestellt. Zu jedem Bildpixel eines Dokumentationsbildes liefert die Funktion als Antwort eine eindeutige Abteilnummer, sofern der Bildpixel zu einem abgetrennten Abteil gehört; gehört der Bildpixel zu keinem Abteil, z. B. weil er außerhalb der abgegrenzten Abteile liegt, wird keine Abteilnummer zurückgeliefert. Die Hinterlegung der Abteilstruktur erfolgt in einer typischen Ausführung der Erfindung manuell bei der Einrichtung des Schubladenschranks. Sie kann optional auch automatisch durch eine Bildanalyse der Leeraufnahmen erfolgen (z. B. durch eine Identifikation der Trennbereiche zwischen den Abteilen mittels kantensensitiven Bildverarbeitungsalgorithmen).
• • (H3) Berechnung von Differenzbildern (Subtraktion zweier Bilder): Die Hilfsfunktion (H3) führt eine pixelweise Subtraktion zweier Bilder gleicher Größe durch (Dokumentationsbilder und/oder Leeraufnahmen). Subtrahiert man beispielsweise das Dokumentationsbild vor einem Entnahmevorgang von dem Dokumentationsbild nach einem Entnahmevorgang, so enthält man nur die Änderungen, die aus dem Entnahmevorgang resultieren. Alle weiten optischen Details der Schubladen bzw. der Abteile in den Schubladen werden ausgeblendet. Um den Einfluss von Schwankungen in der Beleuchtungsstärke zu minimieren, kann optional eine Helligkeitsnormierung an unveränderlichen Bildmerkmalen durchgeführt werden.

The following auxiliary functions can be used in the evaluation procedures:

• • (H1) Blanking: A blank is created for each drawer, defined by the fact that there is no object in the drawer when taking a picture. In addition, the scaling factors of the images are determined (in mm per pixel for each drawer). The recordings are generated in a typical embodiment of the invention in the establishment of the drawer cabinet. The auxiliary function (H1) provides the empty recordings and the scaling factors for the subsequent evaluations.
• • (H2) Depositing the compartment structure (storage surface structure) for each drawer: The auxiliary function (H2) provides the division of each drawer into separate compartments. For each image pixel of a documentation image, the function provides in response a unique compartment number, as long as the image pixel belongs to a separate compartment; the image pixel belongs to no compartment, z. For example, because it is outside the demarcated compartments, no compartment number is returned. The deposit of the compartment structure is done in a typical embodiment of the invention manually when setting up the drawer cabinet. Optionally, it can also be performed automatically by an image analysis of the blank images (for example by an identification of the separation regions between the compartments by means of edge-sensitive image processing algorithms).
• • (H3) Calculation of difference images (subtraction of two images): The auxiliary function (H3) performs a pixel-by-pixel subtraction of two images of the same size (documentation images and / or blank images). If, for example, the documentation image is subtracted from the documentation image after a removal operation before a removal operation, then only the changes resulting from the removal process are included. All the wide optical details of the drawers or the compartments in the drawers are hidden. To minimize the influence of variations in illuminance, brightness normalization can optionally be performed on immutable image features.

If it is not absolutely necessary to detect an incorrect removal immediately (eg no immediate danger to persons) or if erroneous withdrawals are rarely to be expected, a manual evaluation of the documentation screens may be useful. The manual identification can be carried out at any time after a misappropriation becomes known. The evaluation takes place directly on the computer ( 16 ). Optionally, the evaluation can also be done on another computer after the documentation images have been transferred to it.

In a variant of the invention, the identification of the erroneous withdrawals can take place as follows: The stored documentation images are consecutively numbered. If a mismatch has occurred between the two documentation images m1 (eg, drawer cabinet filling time) and m2 (eg, failure detection), the mismatching can be detected by the interval bisecting method. For this purpose, a documentation image m3 is selected by the algorithm, which subdivides the interval m1 to m2 into two approximately equal halves. The user of the system judges, by looking at the image m3, whether the misfire has taken place in the first half (images m1 to m3) or in the second half (images m3 to m2). In this way, the number of removal operations in question is halved. By iterative continuation of this procedure, finally, that documentation image can be identified that documents the misappropriation.

In this way, the date and time (as well as further stored information, if any) are determined for each miscarriage, which constitute an important clue for clarifying the miscarriage.

In the context of the invention, however, an automatic identification of false pickings is possible:
Depending on the area of use of the drawer cabinet, it may be necessary to indicate an incorrect removal immediately after it has occurred, for example, if a specific safety risk is associated with an incorrect removal (removal of a wrong medication package) or if this leads to disruptions in the operation (high number of incorrect withdrawals).

The automatic identification of misfeeds is realized by the execution of image processing algorithms, which are usually applied to the documentation images directly after image acquisition. The image evaluation is done in a typical embodiment of this invention directly on the computer 16 in the drawer cabinet. If the image processing algorithms detect a misreading, this can optionally be indicated to the requester by audible or visual warnings (warning sound, text on a screen, separate visual display). The system can optionally do so that warnings must be acknowledged by the requester. Depending on the area of application, the requestor can optionally be given the opportunity to remedy the incorrect removal. If in this case no correction of the incorrect removal takes place, the drawer cabinet can be locked for further withdrawals. Likewise, the requestor can be prevented from logging out of the system.

Several methods according to the invention for the automatic identification of incorrect pickings are described below. Three different approaches A1 to A3 are presented, which can also be combined. The procedures apply to the following image material: For each individual removal process there are two documentation images, which are designated n1 (background documentation image, image BEFORE the removal process) or n2 (reverse documentation image, image AFTER the removal process). In addition, two so-called object images g1 and g2 are calculated from this: the object image g1 (image-object image, object image BEFORE the removal process) is calculated as a difference image (H3) between n1 and the associated blank image (H1); analogously, the calculation of g2 takes place (return object image, object image AFTER the removal process). In the subject images, the drawer background is hidden, d. H. he appears as a homogeneous monochrome surface.

(A1) Identification of emptied sections:

Spatially separated objects are identified in the object image g1 and their position in the drawer determined (eg by means of the blob analysis technique from image processing known to the person skilled in the art). The auxiliary function (H2) additionally assigns the associated drawer compartment to each object. As a result, there is an object list for the object image g1, in which the individual objects are listed, together with information about the image pixels covered in each case by the object and the assigned compartment. Analogously, a further object list for the object image g2 is calculated.

By comparing the two object lists, it is possible to immediately identify which compartments have been emptied. If only one item is set in each compartment, the number of items removed can also be read directly.

On the basis of the automatic evaluation (A1), it is detected whether one or more sections have been emptied. In addition, it can optionally be checked by comparing with the inventory system whether the compartments have been emptied as intended.

(A2) Image-based identification of objects:

Analogous to the procedure in (A1), first an object list for the object image g1 (image BEFORE the removal process) is calculated. In addition, a separate image file is created for each identified object, which subsequently serves as a recognition mask (so-called template). The image files are extracted from the subject image g1; they contain only the image pixels that belong directly to the object.

The object image g2 (image AFTER the removal process) is then examined to see whether it still contains the identified objects. In this case, in a typical embodiment of this invention, the image processing method "pattern matching" is used, which finds the deposited image files based on correlation (eg based on the image intensities) or feature-based (eg on the basis of edges). The process recognizes objects even after a change of position (translation or rotation). As a result, pattern matching provides the current location of the items. Via the auxiliary function (H2), each identified object is optionally assigned the associated compartment.

) shows a drawer before the removal process (item image g1), ) shows extracted image files as templates and ) shows the drawer after the removal process (object image g2) with marking the objects identified by pattern matching method.

By an analysis of the found objects it can be immediately identified, which objects were taken. It is optionally checked whether the remaining parts of the compartment allocation has changed, so if interchanges have taken place.

Special preconditions for the applicability of this procedure (A2) are: For each object an image file must be generated automatically. This succeeds only when the objects do not touch. This can be ensured by the fact that there is only one object in each compartment or that several objects in a compartment are separated by object holders. Furthermore, the views of objects not removed may only change in translation, rotation and scaling; further changes to the view are not allowed.

(A3) Database-based identification of objects:

On the calculator 16 a database is created in which image files are stored for those items that may be present in the drawer cabinet. If an item in the drawer cabinet can have different views (eg, front and back of a medicine package), multiple views of an item may be stored. If an object has variable properties (eg an expiration date), only partial sections of the views may be stored.

In the two documentation images n1 and n2 and / or in the object images g1 and g2 (images BEFORE and AFTER the removal process), the objects imaged therein are identified. In this case, in a typical embodiment of this invention, the image processing method "pattern matching" is used, which searches for fits for the images stored in the database. Before pattern matching is performed, the image files are adjusted to the scaling factors of the documentation images that are available via the auxiliary function (H1). As a result of pattern matching, the identified items and their exact location in the drawer are available. Via the auxiliary function (H2), each identified object is optionally assigned the associated compartment.

In order to reduce the execution time for pattern matching, it is optionally possible by means of a comparison with the storage system to first determine which objects are in the drawer before the removal operation. Pattern Matching searches only for the image files of these objects.

On the basis of the lists of identified objects before and after the removal process, it is possible to directly determine which objects have been removed. It can be optionally checked whether the remaining parts of the compartment allocation has changed, so if z. B. permutations have taken place. It can optionally be checked whether additional items have been placed in the drawer cabinet (new or refill); Optionally, the stock can be updated in this case.

The prerequisite for the applicability of this method: In contrast to method (A2), this method is also applicable in the case that several objects are in a compartment.

In the context of the present invention, an automatic check of best-before date dates (abbreviated MHD) can also take place:
Items in drawer cabinets may be given a best before date (or expiration date) beyond which use is prohibited (eg, medicines or food). If an item is placed in the drawer cabinet in such a way that the best-before date is visible in the documentation screens, the date can be checked automatically.

For automatic recognition, the documentation image n1 (out-of-documentation image) is analyzed with a text recognition algorithm. A best-before date is identified by two components that are close together: on the one hand, a text marker ("Best before" or similar markings), on the other hand, a date.

If an item with an expired MHD has been identified via this function, this can optionally be indicated to the requester by acoustic or visual warnings (warning sound, on-screen indication, separate visual display). By means of the absolute position of the text component in the documentation image, the best before date can be assigned to a specific object.

Compared to the storage devices and storage and / or removal methods and documentation methods known from the prior art, the device and the method according to the present invention have the following advantages in particular:
With the invention, discrepancies in the stock of drawer cabinets can be traced back to concrete faulty removal operations. If the removal is associated with an authentication process, the person who performed the faulty operation can also be determined.

With the invention, a higher cost-effectiveness in the use of drawer cabinets can be achieved because shortages can be assigned directly to a consumer (if an authentication system is present). It can thus be a billing of costs by organizational units (or other categorizations of the requester).

With the invention, an increased security in dealing with the issued items can be realized, since the requesters are pointed to incorrectly removed items (eg., Wrong drug package). This reduces the likelihood that misappropriated items will be accidentally deployed.

The invention can help requesters remove the required items from a drawer cabinet with increased care and thus reduce the number of erroneous removal operations altogether.

It is also particularly advantageous in the present invention that the image recording unit 10 . 11 can be arranged within a completely closed frame structure or within the drawer cabinet so that only the storage area L of the drawers (sections) can be detected. In other words, the presence of the camera alone and the evaluation of the recorded camera images make it impossible to draw any conclusions about the person who has taken a picking operation. Thus, if no authentication system is desired, the device may also be designed anonymized so that all the requirements of data protection is sufficient without further action.

The fields of application of the present invention are in particular:
Drawer cabinets for consumer goods, such as tools or medicines, drawer-type refrigerators for perishable items, such as biological samples or food and circulation shelves and automatic lift systems for storing items.

Claims (25)

Storage device for storing articles with a frame structure ( 1 ) and at least one relative to the frame structure in two opposite directions (Hinbewegungs- and Rückbewegungsrichtung) movable storage unit ( 2 ) having a storage surface (L) for supporting at least one article thereon, an image acquisition unit ( 10 . 11 ) with which the storage surface of the at least one storage unit at its back and during their return movement relative to the frame structure is at least partially detectable and a connected to the image pickup unit and / or integrally formed with this image evaluation unit ( 16 ) to which image data acquired by the image pickup unit is communicable and which is adapted to generate a backward documentation image from at least one point in time during the forward motion detected by the image pickup unit and to generate a backward documentation image at at least one time during the rewinding by the image pickup unit detected return motion image data. Bearing device according to the preceding claim, characterized in that the position device and / or the image evaluation unit has an evaluation unit for evaluating the back and forth documentation image for changes in the state of the storage surface. Bearing device according to one of the preceding claims characterized in that the storage area of the at least one storage unit in the case of which it is movable back and forth relative to the frame structure can be detected in sections, in that return motion image data is detectable at several points in time during the forward movement by the image acquisition unit and / or at several times during the return movement by the image capture unit, wherein the forward or return motion image data captured at one of the times is respectively imaged onto the image capture unit Section of this storage area correspond, and in that the outward documentation image can be generated by assembling in sequence image motion image data acquired at the multiple points in time during the forward movement and / or the rear documentation image by segmenting return motion image data acquired at the multiple times during the return movement. Bearing device according to the preceding claim, characterized in that With regard to the relative position of the storage unit and the frame structure, the outward documentation image can be assembled in sections only from the forward motion image data that first appears during the forward movement and or with regard to the relative position of the storage unit and the frame structure, the back-documentation image can be assembled in sections only from return-motion image data that last appeared during the return movement and or that the storage area is partially captured in the form of strip-shaped sections perpendicular to the back and / or return movement. Bearing device according to one of the two preceding claims, characterized in that the outward documentation image and / or the back-documentation image from several at different times captured image Sections of the storage area exactly one storage unit can be generated. Bearing device according to one of the preceding claims, characterized by at least one in a predefined, unchangeable position relative to one of the bearing units arranged, with the image pickup unit can be detected and graphically evaluated by the evaluation unit with respect to its relative position to the frame structure position element ( 7 ) and / or by a device for determining the position of at least one of the storage units relative to the position of the frame structure. Bearing device according to the preceding claim, characterized in that the position element is fixedly attached to the one bearing unit and / or that the position element is a bar with position markings and / or that the position element ( 7 ) a plurality of position markers ( 8th ) and / or that the position element ( 7 ) has a plurality of binary coded position markings and / or that the device for determining the position of at least one of the bearing units relative to the position of the frame structure comprises a device for determining a travel of a drive, a spindle drive and / or a rack drive for moving at least one of the storage units and the frame structure is relative to each other. Bearing device according to one of the preceding claims, characterized in that at least one of the relative to the frame structure movable storage units ( 2 ) is a at least partially extractable from the frame structure and retractable into the frame structure storage unit and / or a drawer, the storage surface is at least partially visible when pulling out of the frame structure and when inserted into the frame structure. Bearing device according to one of the preceding claims, characterized by a plurality of relative to the frame structure in each case in the Hinbewegungs- and Rückbewegungsrichtung relative to the frame structure movable storage units ( 2 ) and / or drawers, each having a storage surface for storage of at least one object thereon. Bearing device according to the preceding claim, characterized in that the plurality of storage units ( 2 ) and / or drawers are arranged stacked in a direction perpendicular to the Hinbewegungs- and / or Rückbewegungsrichtung stacked. Bearing device according to one of the preceding claims, characterized in that the storage device is a drawer cabinet, an automatic lift system for storing items or a circulation rack and or in that the frame structure is the housing and / or the frame of a drawer cabinet, an automatic lift system for storing objects or a circulation rack, wherein at least one of the storage units has a frame structure that is movably fixed, at least partially insertable into the frame structure or the frame, and out of the frame structure is again pull-out drawer. Storage device according to the preceding claim, characterized in that the drawer cabinet is a drawer cabinet and / or a refrigerator with drawer system for consumer goods and / or food or a drawer cabinet for consumer goods and / or tools. Bearing device according to one of the preceding claims, characterized in that the storage device and / or the image evaluation unit is a storage and / or evaluation unit for storing the back and / or the rear documentation image and / or for evaluating the back and / or the back-documentation image by means of digital image processing algorithms. Bearing device according to the preceding claim, characterized in that the storage and / or evaluation unit is designed to evaluate the outward and the backward documentation image in respect of objects imaged by the outward documentation image and not imaged by the backward documentation image and or the memory and / or evaluation unit and / or the image evaluation unit are embodied as hardware and / or software within a computer system and / or a PC. Bearing device according to one of the two preceding claims, characterized in that by means of the storage and / or evaluation unit pictorial images of the storage area without on the Storage surface of stored objects of at least one of the storage units can be generated and / or that in the storage and / or evaluation a storage surface structure of the storage surface of at least one of the storage units is stored and / or that by means of the memory and / or evaluation of Hin documentation images, back Documentation images and / or pictorial images of the storage area without stored on the storage area objects differential images are generated. Bearing device according to one of the preceding claims, characterized in that the image pickup unit comprises a camera and / or a CCD or CMOS line scan camera and / or a CCD or CMOS matrix camera. Storage device according to one of the preceding claims, characterized in that the storage device is designed as a multiple drawers comprising drawer cabinet, a plurality of storage units comprehensive automatic lift system for storing items or a multi-storage units comprehensive circulation rack, in which due to mechanical and / or electrical safety devices always only one drawer or only one storage unit can be pulled out. Bearing device according to one of the preceding claims, characterized by a the inside of the frame structure, the surface of at least one of the storage units and / or the position element of at least one of the storage units at least partially illuminating lighting device ( 13 ). Bearing device according to the preceding claim, characterized in that the inside of the frame structure, the surface of at least one of the storage units and / or the position element of at least one of the storage units at least partially illuminating lighting device ( 13 ) is arranged in the interior of the bearing device and / or formed and / or switchable. Bearing device according to one of the preceding claims, characterized in that a lighting device ( 14 ) and a light sensor ( 15 ) are arranged and / or formed such that the light sensor is illuminable or non-illuminable by this lighting device depending on the position of at least one of the storage units relative to the frame structure and that upon a change in its illumination state by the light sensor, the beginning and / or end of Image recordings can be triggered by the image recording unit is / are. Bearing device according to the preceding claim, characterized in that the lighting device ( 14 ) and the light sensor ( 15 ) are arranged and / or formed in the interior of the bearing device. Bearing device according to one of the preceding claims, characterized in that an at least the image pickup unit and connected to the image pickup unit and / or integrally formed with this trained Bildauswerteeinheit device part of the storage device as a retrofit kit for subsequent addition to the frame structure and the at least one storage unit of an existing part of Storage device is formed. Bearing device according to one of the preceding claims, marked by a trained for personal access control to the storage device authentication system and or by a documentation system designed to document inventories of items stored in the storage device on the basis of evaluations of outward and / or backward documentation images. Storage and / or removal method for storing articles in a storage device, wherein a frame structure ( 1 ) and at least one relative to the frame structure in two opposite directions (Hinbewegungs- and Rückbewegungsrichtung) movable storage unit ( 2 ), which has a bearing surface for supporting at least one object thereon, are formed, with an image recording unit ( 10 . 11 ) the bearing surface of at least one bearing unit is detected at least in sections thereof during its forward and backward movement relative to the frame structure, and wherein an image evaluation unit ( 16 ) is connected to the image recording unit and / or formed integrally therewith, whereby image data acquired by the image recording unit is transmitted to the image evaluation unit and wherein the image evaluation unit comprises a forward documentation image at least one time during the forward movement by the image recording unit and a re-documentation image is generated from at least one time during the return movement by the image-capturing unit detected return motion image data. Storage and / or removal method according to the preceding claim, characterized in that a trained according to one of claims 2 to 23 bearing device for carrying out the storage and / or removal method is used.

Images