CAMERA DEVICE AND METHOD FOR DETECTING AN OBJECT

Abstract

A camera device is provided that comprises a camera having an image sensor for recording images of the objects and at least one first deflection element, and wherein the field of view of the camera has at least one first part field of vision with detection of the first deflection element and a second part field of vision with detection of the first deflection element. In this respect, the first deflection element is arranged such that a first perspective of the first part field of vision is a different one than a second perspective of the second part field of vision; the first part field of vision thus provides a different perspective of the object than the first part field of vision so that at least two sides of the object are simultaneously recorded by the image sensor.

Claims

1. A camera device for detecting an object in a stream of objects moved in a longitudinal direction relative to the camera device, wherein the camera device comprises a camera having an image sensor for recording images of the objects and at least one first deflection element, wherein the field of view of the camera has at least one first part field of vision with detection of the first deflection element and a second part field of vision without detection of the first deflection element, and wherein the first deflection element is arranged such that a first perspective of the first part field of vision is a different one than a second perspective of the second part field of vision; the first part field of vision thus provides a different perspective of the object than the first part field of vision so that at least two sides of the object are simultaneously recorded by the image sensor.

2. The camera device in accordance with claim 1, wherein the second perspective is a plan view.

3. The camera device in accordance with claim 1, wherein the first perspective is a side view from the transverse direction transverse to the longitudinal direction.

4. The camera device in accordance with claim 1, that comprises a second deflection element, wherein the field of view of the camera has a third part field of vision with detection of the second deflection element and the second deflection element is arranged such that a third perspective of the third part field of vision is a different one than the first perspective and the second perspective so that three sides of the object are simultaneously recorded by the image sensor.

5. The camera device in accordance with claim 4, wherein the third perspective is a side view from an opposite direction from the first perspective.

6. The camera device in accordance with claim 1, wherein the camera is installed as stationary at a conveying device which conveys the stream of objects in the longitudinal direction.

7. The camera device in accordance with claim 1, that has a third deflection element that is arranged such that it is detected in the second part field of vision.

8. The camera device in accordance with claim 1, that has a fourth deflection element that once again folds the optical reception path of the first perspective folded by the first deflection element.

9. The camera device in accordance with claim 4, that has a fifth deflection element that once again folds the optical reception path of the third perspective folded by the second deflection element.

10. The camera device in accordance with claim 1, wherein the deflection elements are arranged such that the light paths between the camera and the object are of the same length for the different perspectives with a tolerance corresponding to a depth of field range of the camera.

11. The camera device in accordance with claim 1, wherein the deflection elements have a mirror and a holder for installation in a specified arrangement and orientation with respect to the stream of objects.

12. The camera device in accordance with claim 1, wherein the image sensor is configured as a line sensor.

13. The camera device in accordance with claim 1, wherein pixel regions of the image sensor disposed next to one another correspond to the part fields of vision.

14. The camera device in accordance with claim 13, wherein a central pixel region corresponds to the second part field of vision and a lateral pixel region corresponds to further part fields of vision.

15. The camera device in accordance with claim 1, wherein pixel regions of the image sensor disposed above one another correspond to the part fields of vision.

16. The camera device in accordance with claim 1, that has an illumination unit to illuminate the field of view of the camera, in particular the part fields of vision via the respective deflection elements.

17. The camera device in accordance with claim 1, that has a control and evaluation unit that is configured to localize code regions in the image data detected by the image sensor and to read their code content.

18. A method of detecting an object in a stream of objects moved in a longitudinal direction, wherein images of the object are recorded in a field of view and the field of view has a first part field of vision with detection of a first deflection element and a second part field of vision without detection of the first deflection element, and wherein the first deflection element is arranged such that a first perspective of the first part field of vision is a different one than a second perspective of the second part field of vision; the first part field of vision thus provides a different perspective of the object than the second part field of vision so that at least two sides of the object are simultaneously recorded by the image sensor.

Description

[0031] The invention will be explained in more detail in the following also with respect to further features and advantages by way of example with reference to embodiments and to the enclosed drawing. The Figures of the drawing show in:

[0032] FIG. 1: a schematic view of a camera installed at a conveyor belt with objects to be detected;

[0033] FIG. 2 a three-dimensional view of a camera device with folded optical paths for a simultaneous detection from above and from one side;

[0034] FIG. 3 a front view of the camera device in accordance with FIG. 2;

[0035] FIG. 4 a plan view of the camera device in accordance with FIG. 2;

[0036] FIG. 5 a dissection of the light paths in FIG. 2 to explain how the object can be held by light paths of equal length for all the perspectives in the depth of field range;

[0037] FIG. 6 a three-dimensional view of a camera device with folded optical paths for a simultaneous detection from above and from both sides;

[0038] FIG. 7 a front view of the camera device in accordance with FIG. 6;

[0039] FIG. 8 a plan view of the camera device in accordance with FIG. 6;

[0040] FIG. 9 a representation of a conventional camera device that requires three cameras for the detection of three sides; and

[0041] FIG. 10 a representation of a further conventional camera device that detects an object from above with the aid of a mirror with a horizontal orientation.

[0042] FIG. 1 shows a camera 10 that is mounted above a conveyor belt 12 on which objects 14 are conveyed through a field of vision 18 of the camera 10 in a conveying direction 16 indicated by arrows. The objects 14 bear codes 20 that are read by the camera 10 at their outer surfaces in a preferred embodiment. For this purpose, the camera 10 records images of the respective objects 14 located in the field of vision 19 via a reception optics 22 using an image sensor 24.

[0043] An evaluation unit 26 comprises a decoding unit that evaluates the images. In this respect, code regions are identified and the code contents of the codes 20 are read. The evaluation function can also be implemented at least partially outside the camera 10. The camera 10 is only preferably configured as a camera-based code reader. In addition to the reading of optically 1 D or 2D codes, further possible image processing work includes the recognition of symbols, in particular Hazmat labels, the reading of characters (OCR, optical character reading), in particular of addresses, and further processing.

[0044] The camera 10 can be configured as a line scan camera having a linear image sensor 24 of preferably a high resolution of, for example, eight thousand or twelve thousand pixels. Alternatively, the image sensor 24 is a matrix sensor that can have a total comparable resolution overall of four, eight, or twelve megapixels. However, they are distributed over the surface so that a successive image recording with a line in the course of the conveying movement can result in substantially more highly resolved images. In some applications, in particular when using image processing on the basis of machine learning or CNNs (convolutional neural networks), a smaller pixel resolution is also sufficient. A static image recording without a moved object stream or a conveyor belt 12 is generally also conceivable with a matrix sensor. Conversely, it is often sensible also to combine the images recorded by a matrix sensor successively to a larger image in the course of a conveying movement.

[0045] Further sensors that are shown as representative by a feed sensor 28, for example an incremental encoder, by which the speed or the feed of the conveyor belt 12 is determined, can belong to a reading tunnel formed by a camera 10 and a conveyor belt 12. Information that is detected at some point along the conveyor belt 12 can thereby be converted at different positions along the conveyor belt 12, or, which is of equal value thanks to the known feed, at different times. Further conceivable sensors are a trigger light barrier that respectively recognizes the entry of an object 14 into the field of vision 18 or a geometric sensor, in particular a laser scanner, that detects a 3D contour of the objects 14 on the conveyor belt 12.

[0046] The field of vision 18 of the camera 10 is divided by deflection elements 30a-c, in particular mirrors, and the respective optical reception path 32a-b is correspondingly folded. This will become more easily recognizable and will be explained in more detail with reference to FIGS. 2 to 8. Deflection elements 30a, 30b provide that the optical reception path 32a is folded on one side of the object 14. The other optical reception path 32b is deflected by means of a deflection element 30c from a perpendicular extent on the upper side of the object 14 into the horizontal corresponding to the alignment of the camera 10.

[0047] Part fields of vision 18a-b are thereby produced on the surface and on a side surface of the object 14 so that two sides of the object 14 are detectable at the same time. The part fields of vision 18a-b in FIG. 1 are linear for a successive line-wise detection of the object 14 in the course of the conveying movement; with a matrix sensor as the image sensor 24, correspondingly wider part vision fields are produced.

[0048] Different pixel regions or image segments are also produced on the image sensor 24 due to the division of the field of vision 28 into part fields of vision 18a-b. With a line sensor, these image segments are preferably simply disposed next to one another. Accordingly part zones of the reading field not required for the central part of the detection and optionally also their illumination are decoupled and are used for the detection of an additional side by deflections or folding. With a matrix sensor, part fields of vision 18a-b can likewise be arranged next to one another, but also stripwise above one another on the image sensor 24.

[0049] An optional active illumination of the camera 10 not shown in FIG. 1 can likewise be folded via the deflection elements 30a-c as an illumination coaxial to the image sensor 24. A central illumination at the location of the camera or integrated therein is thus sufficient and the respective part fields of vision 18a-b are illuminated thereby.

[0050] The recorded images can be prepared in the evaluation unit 26 or in a downstream image processing using parameters adapted to the part fields of vision 18a-b. Equally, parameters of the image sensor 24 or of the illumination can be set or regulated sectionally. The contrast or brightness is thereby adapted, for example. A beam-shaping or optical filtering, in particular by a corresponding coating, by the deflection elements 30a-c is also conceivable.

[0051] FIG. 2 shows the camera device in accordance with FIG. 1 again in a three-dimensional view. FIGS. 3 and 4 are additionally an associated front view in the conveying direction and a plan view. The camera 10 is aligned horizontally or in parallel with the conveyor belt and is alternatively also aligned against the conveying direction. A lateral optical reception path 32a is guided or folded laterally via an upper deflection element 30a at the left side, first to the bottom next to the conveyor belt 12 and then via a lower deflection element 30b at the left side in as perpendicular a manner as possible onto the side surface of the object 14. The lateral optical reception path 32a corresponds to the part field of vision 18a not separately designated in FIG. 2 for reasons of clarity. The upper deflection element 30a at the left side on its own, that is without a lower deflection element at the left side, could already alternatively deflect to the side surface, but then with an oblique and no longer perpendicular optical path. A central optical reception path 32b is deflected downwardly to the upper side of the object 14 at a central deflection element 30c. The central optical reception path 32b corresponds to the part field of vision 18b no longer separately designated here.

[0052] Thanks to the deflection elements 30a-c and the correspondingly folded optical reception path 32a-b, the upper side and the left side of the object 14 can be simultaneously detected from two different perspectives by the camera 10. It is understood that this would alternatively equally be able to be transferred to the right side. The detection of two sides is anyway admittedly particularly advantageous for the case of a first perspective from the side and of a second perspective from above, but a detection of two other sides or surfaces of the object 14 or two different perspectives than from above and from the side would be equally conceivable.

[0053] FIG. 5 again shows the representation of FIG. 2 as a background and divides the optical reception paths 32a-b therein into their straight-line part sections A-E. The camera 10 with its reception optics 22 only has a limited depth of field range. It is therefore particularly advantageous if the light paths are of equal lengths for the different perspectives. Respective focused images are thereby recorded of the simultaneously detected sides of the object 14. Differences in the light path lengths, particularly when they go beyond the tolerance that a finite depth of field range permits, would in contrast produce blur in the recording of at least one side or surface. The same length of the optical reception paths 32a-b can be ensured by a skillful arrangement of the deflection elements 30a-c, that is the equation A+b=C+D+E could specifically be satisfied. For this purpose, the central deflection element 30c is arranged further away from the camera 10 with respect to the upper deflection element 30a at the left side so that A and B are extended, and indeed by just so much as corresponds to the diversion by the double deflection onto the side surface having the sections C, D, E. It must be repeated that approximate identity within the framework of the depth of field range is sufficient.

[0054] The camera 10 can have an adjustable focus or an autofocus instead of a fixed focus. However, this alone does not solve the problem of blur from different perspectives since the depth of field range can thus only be adapted for one perspective. Another option of solving the focusing problem is a combination with the teaching of EP 2 937 810 A1 named in the introduction. In this respect, deflection elements 30a-c are suitably replaced with staggered deflection elements at different distances. The recorded image sections multiply in accordance with the staggering and a respective image section is localized and further processed that has been recorded with a suitably long light path in the depth of field range.

[0055] FIG. 6 shows a further embodiment of the camera device in a three-dimensional view, with FIGS. 7 and 8 as a supplementary associated front view in the conveying direction or as a plan view. Unlike the embodiment in accordance with FIGS. 2 to 6, the other side of the object 14 is now also detected from an additional third perspective. The above statements apply analogously with respect to the different embodiment options and in particular the configuration of the light paths for a respective recording in the depth of field range. This also in particular applies to a configuration with equally long light paths corresponding to FIG. 5, i.e. the detection of the other side of the object 14 should likewise preferably take place with a light path of equal length.

[0056] To provide a third perspective and to also still detect the second side of the object 14, at the right here observed in the conveying direction 16, an additional decoupling of a further lateral optical reception path 32c takes place laterally via an upper deflection element 30d at the right side, first downward next to the conveyor belt 12 and then via a lower deflection element 30a at the right side in as perpendicular a manner as possible to the other side surface of the object 14. The further lateral optical reception path 32c corresponds to an additional part field of vision 18c only designated in FIG. 7 for reasons of clarity. Two part fields of vision 18a, 18c are now therefore decoupled at both sides of the central part field of vision 18b. Thanks to the deflection elements 30a-e and the correspondingly folded optical reception paths 32a-c, the upper side, the left side, and the right side of the object 14 can be simultaneously detected from three different perspectives by the camera 10. Instead of a detection from above and from both sides, three other perspectives of a different combination of sides of the object 14 would be conceivable.

[0057] If the deflection does not take place in a perpendicular manner on the lateral surfaces, as previously described, but rather within the horizontal plane at a 45° angle, that is, so-to-say on a perpendicular edge of an object 14 imagined as parallelepiped-shaped, the front surface or back surface can also be detected after one another in the course of the conveying movement with the respective side. In the embodiment in accordance with FIGS. 2 to 5, a decision would have to be made on the front surface or the rear surface or to cover the respectively not detected surface via the perspective from above by a corresponding tilting. In the embodiment in accordance with FIGS. 6 to 8, the one side can be detected at the front surface and the other side at the rear surface. However, the disadvantage of a no longer perpendicular perspective on the respective object surface has to be accepted for this purpose.