Machine and method for processing and / or moving metal sheet or metal sheet comprising edge identification device

Abstract

A machine and method for processing or manipulating a sheet metal or sheet metal, the machine comprising a vision system for imaging at least a portion of the sheet or sheet and a processing unit for performing edge recognition. In some embodiments, the step is performed by a deep learning algorithm. Where the identifying step is not satisfactory, the processor processes the image again using a mask device that can be applied to at least a portion of the image to better identify the edge with respect to the background.

Description

Technical Field

[0001] The present invention relates to a machine and method for processing and / or moving sheet metal and / or metal plates, and in particular to a machine for processing and / or moving sheet metal and / or metal plates, semi-finished products, scraps, the machine being provided with an artificial vision system, the artificial vision system being adapted to detect geometric information of a workpiece to be processed, such as position and orientation, and being capable of setting optimal operating parameters for an operating device for processing and / or moving the workpiece. The present invention also relates to a method for processing and / or moving sheet metal and / or metal plates based on geometric information about the workpiece detected by the artificial vision system. Background Art

[0002] Machines for processing and / or moving all or parts of sheet metal and / or metal sheets are known, such as laser cutting machines, punching machines, combined cutting and punching machines, bending machines, transmission and manipulation machines (such as Cartesian manipulators, conveyor belts, etc.), which are provided with an artificial vision system for determining the position and orientation of workpieces or parts to be processed and / or moved, including sheet metal or metal sheets, thereby allowing the control unit of the machine to adjust and configure the operating parameters, i.e. the processing and / or movement program, in order to appropriately control the operating device that performs the processing, such as a cutting device and / or a punching device and / or a bending device or a device for moving or manipulating the workpiece.

[0003] The workpiece to be machined typically comprises sheet metal and / or sheet metal, and sometimes comprises scrap or skeletons of sheet metal or sheet metal that can be used to produce other workpieces.

[0004] Known artificial vision systems for this purpose comprise a camera which is able to take a photograph or an image of the workpiece to be processed and / or moved.

[0005] The processing and computing unit of the artificial vision system processes the image acquired by the camera by means of a suitable contour extraction algorithm to extract and infer the contour of the workpiece in the processed image, that is, the line or a set of lines that limit or delimit the workpiece.

[0006] In some applications, the inferred profile is compared to a previously saved reference map or drawing of the same workpiece to obtain geometric information related to the alignment of the workpiece on the work plane with reference to the machine's reference system.

[0007] The alignment information comprises in particular an offset or distance and a rotation of one or more contour stretches, in particular substantially straight contour stretches, of the workpiece shown in the processed image relative to a corresponding contour stretch of the workpiece shown in the reference image. In this way, the position and orientation of the workpiece relative to a reference system of the machine can be determined. For example, the "position" of a workpiece refers to the distance of a reference point (usually an edge or an angle) of the workpiece from the origin of the machine reference system along two orthogonal axes, while the "orientation" of the workpiece refers to the angle formed by a reference stretch of the workpiece contour with one of the two axes of the reference system.

[0008] The geometric information thus obtained about the position and orientation of the workpiece is sent to the machine's control unit, which uses this information to adjust and configure the working parameters, i.e. the machining program, and then appropriately controls the operating device that performs the work, such as a cutting device and / or a punching device and / or a bending device or a moving and manipulating device.

[0009] A disadvantage of the known processing and / or moving machines and the associated processing and / or moving methods is that the contour extraction algorithm is usually not able to accurately and completely identify and recognize the contour of the workpiece, and in particular is not able to distinguish all contour stretches from the background, i.e. the working plane, which is usually due to low contrast with the background, surface treatment, material type, coloring of the workpiece and / or due to lighting conditions and / or the position of the workpiece relative to the camera (viewing angle), etc.

[0010] Furthermore, the artificial vision systems do not allow the alignment and positioning of scraps or frames of sheet metal or metal plates to be determined without a reference diagram or drawing to compare to.

[0011] Therefore, in both cases, an operator needs to intervene via a control panel and manually select the necessary points on the processed image of the workpiece to allow the artificial vision system to identify missing or underrepresented contour stretches of the workpiece.

[0012] However, this manual identification process, besides requiring the intervention of an experienced operator, is rather laborious and time-consuming and may in any case lead to errors. Summary of the invention

[0013] The object of the invention is to improve the known machines and methods for processing and / or moving sheet metal and / or metal plates.

[0014] Another object is to provide a machine and a method which allow the position and orientation of a workpiece arranged on a working plane of the machine to be determined in a precise, accurate and substantially automatic manner, and then the operating parameters of an operating device which processes and / or moves / manipulates the workpiece to be configured and / or adjusted.

[0015] A further object is to provide a machine and a method which allow determining at least the position and orientation of a workpiece, including sheet metal, metal sheet, semi-finished product, scrap of sheet metal or metal sheet or a frame, on a working plane of the machine, in particular without having to process and save reference images or drawings of the workpiece.

[0016] Another object is to provide a machine and a method which allow geometric information of a workpiece to be machined to be determined, in particular comprising at least the distance and the angle of rotation of one or more contour stretches of the workpiece.

[0017] In a first aspect of the present invention, a machine for processing sheet metal and / or metal plates according to claim 1 is provided.

[0018] In a second aspect of the invention, a method for processing sheet metal and / or metal plate according to claim 8 is provided.

[0019] In a third aspect of the present invention, a machine for processing sheet metal and / or metal plates according to claim 15 is provided.

[0020] In a fourth aspect of the present invention, a method for processing sheet metal and / or metal plate according to claim 22 is provided. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The present invention may be better understood and implemented with reference to the accompanying drawings, which show exemplary and non-limiting embodiments of the present invention, in which:

[0022] - Figure 1 is a partial view of a machine according to the invention for processing and / or moving sheet metal and / or metal plates, the machine being associated with a workpiece to be processed, in particular a semi-finished workpiece, which is arranged on a working plane of the machine;

[0023] - Figure 2 is an image of a workpiece, in particular a new sheet of metal, arranged on a working plane of a machine, said image being obtained by Figure 1 The machine's artificial vision system camera acquires;

[0024] - Figure 3It is the processing unit of the machine that uses deep learning algorithms to Figure 2 a processed image obtained by processing an image;

[0025] - Figure 4 yes Figure 3 an enlarged view of the processed image showing a contour stretch of the workpiece contour;

[0026] - Figure 5 is an image of a further workpiece to be machined which is positioned on the working plane of the machine, the image being obtained by Figure 1 The machine's artificial vision system camera acquires;

[0027] - Figure 6 is a stereogram of a processed image of a workpiece, in particular sheet metal, processed by a contour recognition algorithm;

[0028] - Figure 7 is a stereogram of a processed image processed by a contour recognition algorithm of another workpiece, in particular sheet metal;

[0029] - Figure 8 By straightening Figure 6 a plan view of a processed image obtained by processing an image of

[0030] - Fig. 9 A mask device is shown, the mask device comprising a plurality of elliptical masks, the masks being applied to Figure 8 a portion of the processed image of the image, the portion containing the contour stretch to be identified and recognized, and Fig. 9 showing an enlarged view of one mask of a portion of the mask including a contour stretch;

[0031] - Fig.10 A mask arrangement is shown, the mask arrangement comprising a plurality of elliptical shaped masks which are applied to the film by straightening Figure 7 a portion of the processed image obtained by processing the image of the image, the portion containing the contour stretch to be identified and recognized, and Fig.10 showing an enlarged view of one mask of a portion of the mask including a contour stretch;

[0032] - Figures 11a to 11c shows two elliptical shaped masks enclosing respective portions of the contour stretch and a corresponding enlarged view of a rectangular shaped mask enclosing a portion of the contour stretch, the rectangular shaped mask not being the object of the present invention;

[0033] - Figures 12a to 12d The invention shows a plurality of masks of a mask device, wherein the plurality of masks are arranged along a detection line and are respectively Figure 8The horizontal top and bottom and the vertical left and right sides of the workpiece outline are superimposed on the contour extension. DETAILED DESCRIPTION

[0034] Reference Figure 1 , partially and schematically shows a machine 1 for processing and / or moving sheet metal and / or metal plates according to the present invention, the machine 1 comprising: an operating device 2, which is capable of processing and / or moving workpieces 50, 51, 52 including sheet metal and / or metal plates; an artificial vision system 10, which is provided with one or more cameras 11, which are suitable for acquiring one or more images 20 of one or more workpieces 50, 51, 52 positioned on a working plane 3 of the machine 1; and a control unit 5, which is connected to the operating device 2 and the artificial vision system 10, and is configured to configure and / or adjust operating parameters, in particular, configure and / or adjust operating parameters based on geometric information of the workpiece to control the operating device 2.

[0035] For example, the workpiece comprises a sheet metal 50 or a metal plate, in particular a rectangular-shaped sheet metal 50 or a metal plate, or comprises a semi-finished product 51 (i.e. a workpiece that has been partially processed, such as Figure 1 ), or scrap 52 comprising sheet metal or metal plate (obtained from previous processing, such as Figure 5 ).

[0036] exist Figure 1 In the embodiment shown in , the machine 1 is, for example, a machine for processing, in particular a machine for laser cutting, the workpiece and the operating device 2 include a laser cutting head, which can move above a working plane 3 and along three orthogonal axes XYZ of a reference system S of the machine (the origin of the reference system S is a defined reference point, for example, located on the working plane 3), and the working plane is fixed.

[0037] The artificial vision system 10 comprises one or more, for example two, cameras 11 of known type and commercially available, which are configured to acquire at least one image 20 of one or more workpieces positioned on the working plane 3. The processing unit 12 is arranged to process the acquired image 20.

[0038] The processing unit 12 is connected to the camera 11 and to the control unit 5 in order to transmit to the control unit 5 geometric information relating to one or more workpieces 50 and obtained by processing said image 20 , as better explained in the following description.

[0039] The processing unit 12 can be separate and different from the control unit 5, for example, the processing unit 12 can be associated with the artificial vision system 10, or the processing unit 12 can be embedded in the control unit 5, that is, overlap with the control unit 5, and the control unit 5 is used as the processing unit 12 of the image 20 acquired by the camera 11.

[0040] The processing unit 12 is configured to process the image 20 acquired by the camera 11 by means of a deep learning algorithm to identify and obtain a processed shape image 30 in which a shape 150 of the workpiece 50 with a corresponding contour 160 is shown. In the processed shape image 30, the shape 150 of the workpiece 50 acquired by the deep learning algorithm is superimposed on the image of the workpiece 50 acquired by the camera ( Figure 3 and Figure 4 ).

[0041] “Contour” refers to a line or collection of straight and / or curved lines that limits and circumscribes an area defining a shape 150 , ie, the form in which the workpiece 50 is identified and defined by the deep learning algorithm.

[0042] It should be noted that these contour lines include actual edges and / or corners of the workpiece defining their extension as well as lines dividing areas of the image having different intensities and / or colors.

[0043] Profile 160 includes the outer contour of the workpiece (e.g., Figures 2 to 5 The profile 160 may include two longer straight horizontal stretches or edges 160a and two shorter straight vertical stretches or edges 160b of the sheet metal in the workpiece, but the profile 160 also includes an internal profile that is specific to machined internal areas, such as holes, openings, etc., made in the workpiece (e.g., Figure 1 The opening formed in the semi-finished product 51 and Figure 5 The opening is defined by the waste material 52.

[0044] A deep learning algorithm is a known type of artificial intelligence algorithm, for example Google's DeepLabv3 algorithm, which, after appropriate training steps, is able to recognize and distinguish artifacts of different shapes and sizes and to process corresponding shapes or masks with recognizable contours or edges.

[0045] If the contour 160 of the shape 150 processed by the deep learning algorithm is accurately and completely defined and can be completely (i.e. in all its contour stretches 160a, 160b) distinguished from the background of the processed shape image 30 (i.e. from the working plane 3) (and therefore if the contour 160 can be completely and accurately distinguished and identified), the processing unit 12 is able to distinguish and identify one or more contour stretches 160a, 160b of the contour 160 of the shape 150, and extract geometric information of the workpiece 50 based on the identified contour stretches 160a, 160b.

[0046] As is known, a contour or a contour stretch is defined from the background of the image in an accurate and distinguishable manner when any known type of contour recognition algorithm, such as the Canny algorithm, is able to successfully detect such a contour or stretch correctly and completely in an unambiguous manner, i.e., when any known type of contour recognition algorithm is able to successfully complete the recognition process.

[0047] If, on the other hand, the contour 160 of the shape 150 processed by the deep learning algorithm is not precisely and completely defined (in particular when the deep learning algorithm roughly identifies the area of ​​the image 20 of the workpiece 50 where the material is shown, but does not allow the edge recognition algorithm to accurately identify all contour stretches), and the contour 160 of the shape 150 processed by the deep learning algorithm includes at least one stretch that is indistinguishable or difficult or almost indistinguishable from the background of the processed shape image 30, i.e., the working plane 3 (and therefore the contour 160 cannot be accurately and completely discerned and recognized), the processing unit 12 processes the image 20 previously acquired by the camera 11 with the aid of a known contour recognition algorithm, such as the Canny algorithm, etc., to obtain a processed contour image 40, which includes the contour 260 of the workpiece 50. The “contour” 260 refers to a line or a collection of straight lines and / or curves that limit and define the outer contour of the workpiece 50 and / or the inner contour of a hole, opening, etc. present on the workpiece.

[0048] The processing unit 12 is also configured to process the processed shape image 40 with the aid of a mask device 15, which is applied to at least a portion of the processed shape image 40, wherein the at least a portion includes at least one contour extension 260a of the shape 260, which is not precise enough, i.e., it cannot be distinguished or is difficult to distinguish from the background of the processed shape image 40, i.e., the working plane 3.

[0049] The shape 150 in the processed shape image 30 is used to identify the portion of the processed contour image 40 to which the masking device 15 is to be applied, and in particular, the two processed images of the contour 30 and the shape 40 are superimposed. In other words, using the processed shape image 30 processed by the deep learning algorithm, one or more portions of the processed contour image 40 containing the contour stretch 260a of the contour 260, which is not precise enough and cannot be distinguished from the background, can be identified, and the masking device 15 is to be applied to the contour stretch 260a to process the processed contour image 40 and distinguish and identify the contour stretch 260a of the contour 260.

[0050] Therefore, the mask device 15 advantageously allows the contour stretch 260a of the outer shape 260 bounded by the mask device 15 to be isolated from the image background, i.e., the working plane 3, and in particular allows the contour stretch 260a of the outer shape 260 bounded by the mask device 15 to be isolated from the image background, i.e., the working plane 3 with higher accuracy and clarity, so that the processing unit 12 can accurately distinguish and identify the contour stretch 260a of the outer shape 260.

[0051] In particular, the mask device 15 includes one or more masks 16, in particular a plurality of masks 16, which are arranged to be aligned along the detection lines R, T and regularly spaced, and the plurality of masks can be superimposed on the contour extension 260a of the contour 260, in particular on the contour extension that is not precisely defined and not completely distinguishable, each mask 16 is defined by a closed plane curve, which is in particular elliptical or oval or circular in shape, and the mask 16 surrounds a corresponding portion of the processed contour image 40, and the corresponding portion of the processed contour image 40 contains a corresponding portion of the contour extension 260a.

[0052] like Fig. 9 , Fig.10 , Fig.11a and Fig.11b As is clearly shown in the drawings, a mask 16 is shown in these drawings, which has, for example, an elliptical shape, which particular elliptical shape allows to better isolate the contour stretch 260a of the profile, which is essentially straight, imprecise and / or difficult to distinguish, because it is partial and / or because due to the low contrast of the image there are other "interfering" or "noise" stretches 261, 262, which, for example, belong to the support elements ( Fig. 9 ) and / or the surface reflection and / or surface treatment portion of the workpiece 50 ( Fig.10 ). In particular, the level ( Fig.11b) or tilt ( Fig.11a )'s contour stretch 260a can also be clearly and unambiguously discerned and identified, even when there are multiple "noise" stretches on both sides.

[0053] Fig.11c Also shown is a rectangular mask which is not the object of the invention but which incorporates at least two straight stretches which, for example, form the two upper edges of the mask and can thus be exchanged with the actually searched contour stretches.

[0054] A further advantage of using a mask 16 having an elliptical shape is that a shorter time is required to process only the portion of the processed topography image 40 bounded by the mask 16 having an elliptical shape than the time required to process the entire processed topography image 40 .

[0055] The mask 16 of the mask device 15 may be arranged along the horizontal detection line R and superimposed on the lower horizontal contour stretch 260a and the upper horizontal contour stretch 260b of the outer shape 260 of the workpiece 50 ( Fig.12a , Figure 12b ), and the mask 16 of the mask device 15 may be arranged along the vertical detection line T and superimposed on the left vertical contour extension 260c and the right vertical contour extension 260d of the outline 260 of the workpiece 50 ( Fig.12c , Fig.12d ).

[0056] The processing unit 12 is then able to extract geometric information about the workpiece 50 with reference to the reference system S of the machine 1 , based on one or more contour stretches 260 a of the contour 260 identified and recognized in the processed contour image 40 by means of the mask device 15 .

[0057] The geometric information comprises at least the distance and the rotation angle relative to the reference system S of one or more of said contour stretches 160a, 160b, 260a of the contour 160 or the outline 260 of the shape 150 discerned and identified in the processed image 30, 40 of the workpiece.

[0058] The geometric information may also include the dimensions of one or more contour stretches, the shape and dimensions of the entire workpiece, and / or the position, orientation, shape and dimensions of any internal areas of the machined (semi-finished) workpiece, in particular openings, holes, grooves, etc.

[0059] The processing unit 12 is configured to send the geometric information of the workpiece thus obtained to the control unit 5 of the machine 1 , in order to configure and / or adjust the operating parameters of the operating device 2 .

[0060] In particular, the geometric information extracted from the artificial vision system 10 allows the control unit 5 to calculate precisely the position and orientation of each workpiece 50 arranged on the work plane 3 relative to the reference system S of the machine 1. The "position" of a workpiece refers, for example, to the distance of a reference point of the workpiece from the origin of the reference system S of the machine along two orthogonal axes X, Y, while the "orientation" of a workpiece refers to the angle formed by a reference stretch of the contour of the workpiece with one of the two axes of the reference system.

[0061] The operation of the machine 1 can be described by the various steps described below defining the method of the present invention for processing sheet metal and metal plates.

[0062] The method according to the invention for processing sheet metal and / or metal sheet in a machine 1 provided with a handling device 2 for processing and / or moving workpieces 50, 51, 52 comprising sheet metal and / or metal sheet, an artificial vision system 10 for acquiring an image 20 of at least one workpiece 50, 51, 52, and a control unit 5 for controlling the handling device 2 and connected to the artificial vision system 10 comprises the following steps:

[0063] - positioning at least one workpiece 50 , 51 , 52 to be processed on the working plane 3 of the machine 1 ;

[0064] - acquiring at least one image 20 of the workpiece 50 , 51 , 52 using at least one camera 11 of the artificial vision system 10 ;

[0065] - processing the image 20 by means of a deep learning algorithm in order to identify and obtain a processed shape image 30 containing the shape 150 of the workpiece 50 , 51 , 52 with a corresponding contour 160 ;

[0066] - if the contour 160 of the shape 150 in the processed shape image 30 is accurately and completely defined and can be completely distinguished from the background of the processed shape image 30, i.e. the working plane 3, at least one contour stretch 160a, 160b of the contour 160 is discerned and recognized; or

[0067] - if the outline 160 of the shape 150 in the processed shape image 30 is not precisely and completely defined and includes at least one outline stretch that is indistinguishable (i.e., barely or barely distinguishable) from the background of the processed shape image 30, then

[0068] - processing the image 20 by means of a contour recognition algorithm and obtaining a processed contour image 40 comprising the contour 260 of the workpiece 50, 51, 52, and

[0069] - processing the processed contour image 40 by means of a masking device 15, the masking device 15 being applied to at least a portion of the processed contour image 40, said at least a portion comprising at least one contour stretch 260a, 260b, 260c, 260d of the contour 260, said at least one contour stretch being indistinguishable in particular from the background of said processed contour image 40 and said at least a portion being identified on the basis of the shape 150 in the processed shape image 30, in particular by superimposing the processed shape image 30 with the processed contour image 40, in order to precisely isolate the contour stretch 260a of the contour 260 delimited by the masking device 15 from the background, thereby distinguishing and identifying the contour stretches 260a, 260b, 260c, 260d of the contour 260;

[0070] - extracting geometric information of the workpiece 50, 51, 52 based at least on the contour stretches 160a, 160b, 260a, 260b, 260c, 260d of the contour 160 or contour 260, which are identified and recognized in the processed shape image 30 or processed contour image 40 by means of the mask device 15 and with reference to the reference system S of the machine 1;

[0071] The geometric information is sent to the control unit 5 and the operating parameters of the operating device 2 are configured and / or adjusted.

[0072] In particular, the contour 160 and / or the profile 260 of the shape 150 with respect to the workpiece 50 comprises a respective plurality of contour stretches 160a, 160b, 260a, 260b, 260c, 260d, which in particular are substantially rectilinear.

[0073] The geometric information comprises the distances and rotation angles of one or more profile stretches 160a , 160b , 260a , 260b , 260c , 260d of the profile 160 or of the outline 260 relative to a reference system S of the machine 1 .

[0074] The method further comprises calculating the position and orientation of the workpiece 50 positioned on the working plane 3 relative to the reference system S based on geometric information extracted from the analysis of the contour of the workpiece identified and recognized in the processed shape image 30 or in the processed contour image 40 by means of the mask device 15 .

[0075] The mask device 15 used particularly includes at least one mask 16, which is formed by a closed plane curve, which is particularly elliptical or oval or circular in shape, and which surrounds the following part of the processed shape image 40: the part contains the corresponding part of the contour stretch 260a, 260b, 260c, 260d to be distinguished and identified.

[0076] In particular, the mask device 15 includes a plurality of masks 16, each of which has an elliptical, oval or circular shape, and the plurality of masks 16 are arranged to be aligned along the detection lines R, T and regularly spaced apart, and the plurality of masks 16 can be superimposed on a defined contour extension portion 260a of the outer shape 260.

[0077] According to the method of the present invention, it is also provided to acquire the image 20 of the workpiece 50 by superimposing two partial images acquired by the corresponding two cameras 11 of the artificial vision system 10 .

[0078] The workpiece to be processed, which is positioned on the working plane 3 , comprises sheet metal 50 , sheet metal, semi-finished workpiece 51 , or scrap 52 of sheet metal or sheet metal.

[0079] Due to the machine and method for processing sheet metal and / or metal plates of the present invention, by using an artificial vision system 10 provided with a camera 11 and a processing unit 12 (for example, the same control unit 5 of the machine 1), the geometric information of the workpiece positioned on the working plane 3 of the machine 1 can be determined in a precise, accurate and automatic manner without the intervention of an operator, in particular, the position and orientation of the workpiece positioned on the working plane 3 of the machine 1 relative to the reference system S of the machine 1 can be determined, and based on the geometric information, the operating parameters of the operating device 2 that processes and / or moves the workpieces 50, 51, 52 can be configured and / or adjusted.

[0080] In particular, by using a deep learning algorithm, the image 20 of each workpiece 50 acquired by the camera 11 can be processed, and a processed shape image 30 can be obtained, which includes the shape 150 of the workpiece 50, from which the contour 160 is directly extracted, i.e., identified and recognized, and then the geometric information of the workpiece 50 (the distance and rotation angle of each contour extension 160a, 160b of the contour 160 relative to the reference system S of the machine 1) is extracted to configure and / or adjust the operating parameters of the operating device 2 without manual intervention by the operator.

[0081] It is worth noting that the time required for the deep learning algorithm to process the image and extract the contour 160 of the workpiece is less than the time required for a known type of contour recognition algorithm to distinguish and recognize the contour in the image.

[0082] Alternatively, if the contour 160 of the shape 150 of the workpiece 50 in the processed shape image 30 obtained by the deep learning algorithm is not sufficiently defined and is not clear enough to directly identify and distinguish the contour 160, then the contour 160 and the corresponding shape 150 can be advantageously used to identify a portion of the processed shape image 40 (the processed shape image 40 is obtained by processing the image 20 using the contour recognition algorithm), and the mask device 15 is applied to the portion to better distinguish and identify the contour extension 260a of the shape 260 of the workpiece 50 contained in the said portion of the processed shape image 40, and in this case, no manual intervention by the operator is required.

[0083] The position and orientation of the workpiece on the working plane 3 of the machine can also be determined, since no corresponding drawings or reference figures are required to identify and recognize the contour of the workpiece, which includes sheet metal 50, metal sheet, semi-finished product 51, scrap of sheet metal or metal sheet or frame 52.

[0084] The internal contour of the workpiece, which defines already machined internal areas such as holes, openings, etc., can also be determined by deep learning algorithms.

[0085] A variant of a machine for processing sheet metal and / or metal plates is provided, which variant differs from the above-mentioned embodiment in that the processing unit 12 is configured to process an image 20 acquired by at least one camera 11 with the aid of a contour recognition algorithm to obtain a processed shape image 40, which includes the shape 260 of the workpiece 50.

[0086] The processing unit 12 is further configured to process the processed contour image 40 by means of a masking device 15, which is applied to at least a portion of the processed contour image 40, said at least a portion comprising at least one contour stretch 260a of the contour 260, in particular said at least one contour stretch 260a is substantially rectilinear, said at least one contour stretch 260a being indistinguishable or difficult to distinguish from the background of the processed contour image 40, i.e. the working plane 3. The masking device 15 comprises at least one mask 16 defined by a closed plane curve, in particular of elliptical or oval or circular shape, which surrounds the portion of the processed contour image 40 comprising the contour stretch 260a, so as to isolate said contour stretch 260a from the background of the processed contour image 40 and to highlight said contour stretch 260a.

[0087] The processing unit 12 is also configured to: identify and recognize the contour extension 260a of the shape 260; extract geometric information of the workpiece 50 based at least on the contour extension 260a of the shape 260 identified and recognized in the processed shape image 40 and with reference to the reference system S of the machine 1; and send the geometric information to the control unit 5 to configure and / or adjust the operating parameters of the operating device 2, which is arranged to perform at least one predetermined processing and / or movement on the workpiece 50.

[0088] The processing unit 12 is also configured to: identify the portion of the processed shape image 40 to which the mask device 15 is to be applied based on instructions manually provided by an operator, or based on a shape 150 of the workpiece 50 having a corresponding contour 160 (the shape 150 is shown in the processed shape image 30 obtained by processing the image 20 with the aid of an image recognition algorithm).

[0089] In particular, the operator can indicate the portion of the processed shape image 40 to which the mask device 15 is to be applied by selecting one or more areas of the screen on a screen of a control panel of the machine, the control panel being connected to the control unit 5 and the processing unit 12, on which the processed shape image 40 is displayed, and the one or more areas correspond to the portions to be identified of the processed shape image 40.

[0090] Alternatively, a deep learning algorithm may be used as an image recognition algorithm to process the image 20 acquired by the camera 11 and recognize and obtain a processed shape image 30 containing the shape 150 of the workpiece 50 with the corresponding contour 160 .

[0091] The operation of this variant of the machine 1 of the invention can be described by the various steps described below defining the method of the invention for processing and / or moving sheet metal and metal plates.

[0092] The method according to the invention for processing and / or moving sheet metal and / or metal plates in a machine 1 provided with a handling device 2 for processing and / or moving workpieces 50, 51, 52 comprising sheet metal and / or metal plates, an artificial vision system 10 for acquiring an image 20 of at least one workpiece 50, 51, 52, and a control unit 5 for controlling the handling device 2 and connected to the artificial vision system 10 comprises the following steps:

[0093] - positioning at least one workpiece 50 , 51 , 52 to be processed on the working plane 3 of the machine 1 ;

[0094] - acquiring at least one image 20 of the workpiece 50 , 51 , 52 using at least one camera 11 of the artificial vision system 10 ;

[0095] Processing the image 20 by means of a contour recognition algorithm and obtaining a processed contour image 40 , which comprises the contour 260 of the workpiece 50 , 51 , 52 ;

[0096] - processing the processed contour image 40 by means of a masking device 15, the masking device 15 being applied to at least a portion of the processed contour image 40, the at least a portion comprising at least one contour stretch 260a of the contour 260, in particular the contour stretch 260a being substantially rectilinear, the contour stretch 260a being in particular indistinguishable from the background of the processed contour image 40, the masking device 15 comprising at least one mask 16 defined by a closed plane curve, in particular of elliptical or oval or circular shape, the mask 16 enclosing the portion of the processed contour image 40 comprising the contour stretch 260a, so as to isolate the contour stretch 260a from the background of the processed contour image 40, i.e. the working plane 3, and to highlight the contour stretch 260a;

[0097] - Identifying and recognizing the contour extension 260a of the contour 260;

[0098] - extracting geometric information of the workpieces 50 , 51 , 52 based at least on the contour stretch 260 a of the contour 260 identified and recognized in the processed contour image 40 and with reference to the reference system S of the machine 1 ;

[0099] - Sending the geometric information to the control unit 5 for configuring and / or adjusting operating parameters of the operating device 2 .

[0100] The contour 260 includes multiple contour stretches 260a, 260b, 260c, 260d, in particular, the multiple contour stretches are basically straight, and processing the processed contour image 40 by the mask device 15 includes applying the mask device 15 to each of the multiple contour stretches 260a, 260b, 260c, 260d.

[0101] The mask device 15 includes a plurality of masks 16, each of the plurality of masks 16 is defined by a closed plane curve, in particular, the closed plane curve is elliptical, oval or circular in shape, the plurality of masks 16 are arranged to be aligned along the detection lines R, T and regularly spaced apart, and the plurality of masks 16 can be superimposed on the contour extension portions 260a, 260b, 260c, 260d of the outer shape 260.

[0102] The geometric information of the workpiece 50 includes the distance and the rotation angle of one or more contour stretches 260 a , 260 b , 260 c , 260 d of the contour 260 relative to the reference system S.

[0103] The method further comprises calculating the position and orientation of the workpiece 50 positioned on the working plane 3 relative to the reference system S based on geometric information extracted from the analysis of the contours of the workpiece identified and recognized in the processed topographic image 40 by means of the mask device 15 .

[0104] The method includes: identifying the part of the processed shape image 40 to which the mask device 15 is to be applied based on instructions manually provided by an operator, or based on the shape 150 of the workpiece 50 having a corresponding contour 160 and shown in the processed shape image 30 obtained by processing the image 20 with the aid of an image recognition algorithm.

[0105] Advantageously, the image recognition algorithm is a deep learning algorithm adapted to process the image 20 and to recognize and obtain a processed shape image 30 containing the shape 150 of the workpiece 50 with a corresponding contour 160 .

[0106] According to the method of the present invention, it is also provided to acquire the image 20 of the workpiece 50 by superimposing two partial images acquired by the corresponding two cameras 11 of the artificial vision system 10 .

[0107] The workpiece to be processed, which is positioned on the working plane, comprises sheet metal 50 , sheet metal, semi-finished product 51 , or scrap 52 of sheet metal or sheet metal.

[0108] By means of variations of the machine and method for processing sheet metal and / or metal plates according to the present invention, the geometric information of workpieces 50, 51, 52 positioned on the working plane 3 of the machine 1 can be accurately determined, in particular the position and orientation of the workpieces 50, 51, 52 relative to the reference system S of the machine 1 can be accurately determined, and based on this geometric information the operating parameters of the operating device 2 that processes and / or moves the workpiece can be configured and / or adjusted.

[0109] In particular, the mask device 15 comprises a plurality of masks 16 which are elliptical in shape and arranged to be aligned and regularly spaced along the detection lines R, T, the plurality of masks 16 being applied to a portion of a processed contour image 40 obtained by processing the image 20 acquired by the camera 11 by means of a contour recognition algorithm, the mask device 15 allowing accurate and rapid identification and recognition of a contour stretch 260a of the contour 260, even if the contour stretch 260a of the contour 260 is not sufficiently defined and is difficult to distinguish from the background (e.g. Figure 10 to Figure 1 2). Based on instructions provided manually by an operator acting on a machine control panel or advantageously based on a shape 150 of a workpiece 50, one or more portions to which the mask 16 is applied may be identified in the processed shape image 40, which is obtained by processing the same image 20 of the workpiece 50 by means of an image recognition algorithm, in particular a deep learning algorithm.

[0110] Based on the contour stretches 260a, 260b, 260c, 260d of the contour 260 identified and recognized in this way in the processed contour image 40, precise and accurate geometric information of the workpiece 50 can be extracted, allowing the position and orientation of the workpiece 50 relative to the reference system S on the working plane 3 to be calculated.

[0111] Thus, the position and orientation of the workpiece on the working plane 3 of the machine 1 can be determined, since no corresponding drawings or reference diagrams are required to discern and identify the contour of the workpiece, which workpiece comprises sheet metal 50, sheet metal, semi-finished product 51, scrap of sheet metal or sheet metal or a frame 52.

Claims

1. A machine (1) for processing and / or moving sheet metal and / or metal plates, the machine (1) comprising: - a handling device (2) for processing and / or moving workpieces (50, 51, 52) comprising sheet metal and / or metal plates; - an artificial vision system (10) provided with at least one camera (11) capable of acquiring at least one image (20) of at least one workpiece (50, 51, 52) positioned on a working surface (3) of the machine (1); as well as - a control unit (5) for controlling the operating device (2) and connected to the artificial vision system (10); characterized in that the machine (1) comprises a processing unit (12) configured to: - processing the image (20) by means of a deep learning algorithm to identify and obtain a processed shape image (30), the processed shape image (30) comprising a shape (150) of the workpiece (50, 51, 52), the shape (150) having a corresponding contour (160); - if the contour (160) is accurately and completely marked and can be completely distinguished from the background of the processed shape image (30), at least one contour stretch (160a, 160b) of the contour (160) is identified and recognized; or - if the contour (160) of the shape (150) is not accurately and completely marked and the contour (160) of the shape (150) comprises at least a contour stretch that is indistinguishable from the background of the processed shape image (30), then - processing the image (20) by means of a contour recognition algorithm and obtaining a processed contour image (40), the processed contour image (40) comprising the contour (260) of the workpiece (50, 51, 52), and - processing the processed shape image (40) by means of a masking device (15), the masking device (15) being applied to at least one portion of the processed shape image (40), the portion of the processed shape image (40) comprising at least one contour stretch (260a, 260b, 260c, 260d) of the shape (260), in particular the contour stretch (260a, 260b, 260c, 260d) being indistinguishable from a background of the processed shape image (40), the processed shape being masked based on the shape (150) in the processed shape image (30); the portion of the processed shape image (40), in particular, identifying the portion of the processed shape image (40) based on the shape (150) in the processed shape image (30) by superimposing the processed shape image (30) and the processed shape image (40) to accurately isolate the contour stretch (260a, 260b, 260c, 260d) of the shape (260) bounded by the mask device (15) from the background, thereby distinguishing and identifying the contour stretch (260a, 260b, 260c, 260d) of the shape (260); - extracting geometric information of the workpiece (50, 51, 52) based on the contour (160) or at least one contour stretch (160a, 160b, 260a, 260b) of the contour (260) identified and recognized in the processed shape image (30) or in the processed shape image (40), and with reference to a reference system (S) of the machine (1); - sending the geometric information to the control unit (5) in order to configure and / or adjust the operating parameters of the operating device (2).

2. The machine (1) according to claim 1, wherein: The geometric information includes at least the distance and rotation angle of one or more contour stretches (160a, 160b, 260a, 260b, 260c, 260d) of the contour (160) or the contour (260) identified and recognized in the processed shape image (30) or in the processed outline image (40) relative to the reference system (S).

3. The machine (1) according to claim 1 or 2, wherein: The control unit (5) is configured to calculate, based on the geometric information, at least the position and orientation of at least one of the workpieces (50, 51, 52) on the working plane (3) relative to the reference system (S).

4. A machine (1) according to any preceding claim, wherein: The mask device (15) comprises at least one mask (16), wherein the mask (16) is formed by a closed plane curve, in particular, the closed plane curve is elliptical, oval or circular in shape, and the mask (16) surrounds the portion of the processed shape image (40), and the portion of the processed shape image (40) includes the corresponding portion of the contour extension portion (260a, 260b, 260c, 260d).

5. A machine (1) according to any preceding claim, wherein: The mask device (15) comprises a plurality of masks (16), each of the masks (16) being formed by a closed plane curve, in particular, the closed plane curve being elliptical, oval or circular in shape, the masks being arranged to be aligned along the detection lines (R, T) and regularly spaced apart, and the masks being capable of being superimposed on the contour extensions (260a, 260b, 260c, 260d) of the outer shape (260).

6. Machine (1) according to any of the preceding claims, wherein: The operating device (2) comprises one or more of a cutting device, a punching device, a bending and / or forming device, a moving and / or manipulating device.

7. Machine (1) according to any of the preceding claims, wherein: The processing unit (12) is included in the control unit (5).

8. A method for processing and / or moving sheet metal and / or metal plates in a machine (1), the machine (1) being provided with an operating device (2) for processing and / or moving workpieces (50, 51, 52) comprising sheet metal and / or metal plates, an artificial vision system (10) for acquiring an image (20) of at least one workpiece (50, 51, 52), and a control unit (5) for controlling the operating device (2) and connected to the artificial vision system (10), the method comprising the following steps: - positioning at least one workpiece (50, 51, 52) to be machined on a working surface (3) of the machine (1); - using at least one camera (11) of the artificial vision system (10) to acquire at least one image (20) of the workpiece (50, 51, 52); - processing the image (20) by means of a deep learning algorithm to identify and obtain a processed shape image (30), the processed shape image (30) comprising a shape (150) of the workpiece (50, 51, 52), the shape (150) having a corresponding contour (160); - if the contour (160) is accurately and completely marked and can be completely distinguished from the background of the processed shape image (30), at least one contour stretch (160a, 160b) of the contour (160) is identified and recognized; or - if the contour (160) of the shape (150) is not accurately and completely marked and the contour (160) of the shape (150) comprises at least a contour stretch that is indistinguishable from the background of the processed shape image (30), then - processing the image (20) by means of a contour recognition algorithm and obtaining a processed contour image (40), the processed contour image (40) comprising the contour (260) of the workpiece (50, 51, 52), and - processing the processed shape image (40) by means of a masking device (15), the masking device (15) being applied to at least one portion of the processed shape image (40), the portion of the processed shape image (40) comprising at least one contour stretch (260a, 260b, 260c, 260d) of the shape (260), in particular the contour stretch (260a, 260b, 260c, 260d) of the shape (260) being indistinguishable from the background of the processed shape image (40), and the processed shape image (30) being masked based on the shape (150) in the processed shape image (30); the portion of the processed shape image (40), in particular, by superimposing the processed shape image (30) and the processed shape image (40) to identify the portion of the processed shape image (40) based on the shape (150) in the processed shape image (30) so as to accurately isolate the contour stretch (260a, 260b, 260c, 260d) of the shape (260) bounded by the mask device (15) from the background, thereby distinguishing and identifying the contour stretch (260a, 260b, 260c, 260d) of the shape (260); - extracting geometric information of the workpiece (50, 51, 52) based on the contour (160) or at least the contour stretch (160a, 160b, 260a, 260b, 260c, 260d) of the contour (260) identified and recognized in the processed shape image (30) or in the processed shape image (40), and with reference to a reference system (S) of the machine (1); - sending the geometric information to the control unit (5) and configuring and / or adjusting the operating parameters of the operating device (2).

9. The method according to claim 8, wherein: The profile (160) and / or the outer shape (260) comprises a plurality of profile stretches (160a, 160b, 260a, 260b, 260c, 260d).

10. The method according to claim 8 or 9, wherein: The geometric information comprises at least the distance and the rotation angle of one or more profile stretches (160a, 160b, 260a, 260b, 260c, 260d) of the profile (160) or of the contour (260) relative to the reference system (S) of the machine (1).

11. The method according to any one of claims 8 to 10, comprising: Based on the geometric information, at least the position and orientation of at least one of the workpieces (50, 51, 52) on the working plane (3) relative to the reference system (S) are calculated.

12. The method according to any one of claims 8 to 11, wherein: The mask device (15) comprises at least one mask (16), wherein the mask (16) is formed by a closed plane curve, in particular, the closed plane curve is elliptical, oval or circular in shape, and the mask (16) surrounds the portion of the processed shape image (40), and the portion of the processed shape image (40) contains the corresponding portion of the contour extension portion (260a).

13. The method according to any one of claims 8 to 12, wherein: The mask device (15) comprises a plurality of masks (16), in particular, each of the masks (16) has an elliptical or oval or circular shape, the masks are arranged to be aligned along the detection lines (R, T) and regularly spaced apart, and the masks can be superimposed on the contour extension (260a, 260b, 260c, 260d) of the outer shape (260).

14. The method according to any one of claims 8 to 13, wherein: The workpiece (50, 51, 52) to be processed comprises one of sheet metal (50), sheet metal, semi-finished product (51), sheet metal scrap (52) or sheet metal scrap.

15. A machine (1) for processing and / or moving sheet metal and / or metal plates, the machine (1) comprising: - a handling device (2) for processing and / or moving workpieces (50, 51, 52) comprising sheet metal and / or metal plates; - an artificial vision system (10) provided with at least one camera (11) capable of acquiring at least one image (20) of at least one workpiece (50, 51, 52) positioned on a working surface (3) of the machine (1); as well as - a control unit (5) for controlling the operating device (2) and connected to the artificial vision system (10); characterized in that the machine (1) comprises a processing unit (12) configured to: - processing the image (20) by means of a contour recognition algorithm to obtain a processed shape image (40), the processed shape image (40) comprising the shape (260) of the workpiece (50, 51, 52); - processing the processed shape image (40) by means of a masking device (15), the masking device (15) being applied to at least one portion of the processed shape image (40), the portion of the processed shape image (40) comprising at least one contour stretch (260a, 260b, 260c, 260d) of the shape (260), in particular the contour stretch (260a, 260b, 260c, 260d) being indistinguishable from the background of the processed shape image (40), the masking device (15) comprising at least one mask (16), the mask (16) being formed by a closed plane curve, in particular the closed plane curve being elliptical or oval or circular in shape, the mask (16) surrounding the portion of the processed shape image (40) so as to isolate the contour stretch (260a, 260b, 260c, 260d) from the background and highlight the contour stretch; - distinguishing and identifying the contour stretches (260a, 260b, 260c, 260d) of the profile (260); - extracting geometric information of the workpiece (50, 51, 52) based on at least the contour stretches (260a, 260b, 260c, 260d) of the shape (260) identified and recognized in the processed shape image (40), and with reference to a reference system (S) of the machine (1); - sending the geometric information to the control unit (5) in order to configure and / or adjust the operating parameters of the operating device (2).

16. The machine (1) according to claim 15, wherein: The geometric information comprises at least the distance and rotation angle of one or more contour stretches (260a, 260b, 260c, 260d) of the shape (260) identified and distinguished in the processed shape image (40) relative to the reference system (S).

17. Machine (1) according to claim 15 or 16, wherein: The control unit (5) is configured to calculate, based on the geometric information, at least the position and orientation of at least one of the workpieces (50, 51, 52) on the working plane (3) relative to the reference system (S).

18. Machine (1) according to any one of claims 15 to 17, wherein: The mask device (15) comprises a plurality of masks (16), each of the masks (16) being formed by a closed plane curve, in particular, the closed plane curve being elliptical, oval or circular in shape, the masks being arranged to be aligned along the detection lines (R, T) and regularly spaced apart, and the masks being capable of being superimposed on the contour extension (260a) of the outer shape (260).

19. Machine (1) according to any one of claims 15 to 18, wherein: The processing unit (12) is configured to: identify at least one of the parts of the processed shape image (40) to which the mask device (15) is to be applied, based on instructions manually provided by an operator, or based on a shape (150) of the workpiece (50, 51, 52) having a corresponding contour (160), wherein the shape (150) of the workpiece (50, 51, 52) is shown in a processed shape image (30) obtained by processing the image (20) with the aid of an image recognition algorithm, in particular, the image recognition algorithm is a deep learning algorithm.

20. Machine (1) according to any one of claims 15 to 19, wherein: The operating device (2) comprises one or more of a cutting device, a punching device, a bending and / or forming device, a moving and / or manipulating device.

21. Machine (1) according to any one of claims 15 to 20, wherein: The processing unit (12) is included in the control unit (5).

22. A method for processing and / or moving sheet metal and / or metal plates in a machine (1), the machine (1) being provided with an operating device (2) for processing and / or moving workpieces (50, 51, 52) comprising sheet metal and / or metal plates, an artificial vision system (10) for acquiring an image (20) of at least one workpiece (50, 51, 52), and a control unit (5) for controlling the operating device (2) and connected to the artificial vision system (10), the method comprising the following steps: - positioning at least one workpiece (50, 51, 52) to be machined on a working surface (3) of the machine (1); - using at least one camera (11) of the artificial vision system (10) to acquire at least one image (20) of the workpiece (50, 51, 52); - processing the image (20) by means of a contour recognition algorithm and obtaining a processed shape image (40), wherein the processed shape image (40) includes the shape (260) of the workpiece (50, 51, 52); - processing the processed shape image (40) by means of a masking device (15), the masking device (15) being applied to at least one portion of the processed shape image (40), the portion of the processed shape image (40) comprising at least one contour stretch (260a, 260b, 260c, 260d) of the shape (260), in particular, the contour stretch (260a, 260b, 260c, 260d) being in contact with the processed shape image (40); The mask device (15) comprises at least one mask (16), the mask (16) being formed by a closed plane curve, in particular, the closed plane curve being elliptical, oval or circular in shape, the mask (16) surrounding the portion of the processed shape image (40) to isolate the contour stretch (260a, 260b, 260c, 260d) from the background of the processed shape image (40) and highlighting the contour stretch; - distinguishing and identifying the contour extension (260a) of the outer shape (260); - extracting geometric information of the workpiece (50, 51, 52) based on at least the contour stretches (260a, 260b, 260c, 260d) of the shape (260) identified and recognized in the processed shape image (40), and with reference to a reference system (S) of the machine (1); - sending the geometric information to the control unit (5) in order to configure and / or adjust the operating parameters of the operating device (2).

23. The method according to claim 22, wherein: The contour (260) includes a plurality of contour stretches (260a, 260b, 260c, 260d), and the processing of the processed contour image (40) by the mask device (15) includes applying the mask device (15) to each contour stretch of the plurality of contour stretches (260a, 260b, 260c, 260d).

24. The method according to claim 22 or 23, wherein: The mask device (15) comprises a plurality of masks (16), each of the masks (16) being formed by a closed plane curve, in particular, the closed plane curve being elliptical, oval or circular in shape, the masks being arranged to be aligned along the detection lines (R, T) and regularly spaced apart, and the masks being capable of being superimposed on the contour extension (260a) of the outer shape (260).

25. The method according to any one of claims 22 to 24, wherein: The geometric information comprises at least a distance and a rotation angle of one or more contour stretches (260a, 260b) of the contour (260) relative to the reference system (S).

26. The method according to any one of claims 22 to 25, comprising: Based on the geometric information, at least the position and orientation of at least one of the workpieces (50) on the working plane (3) relative to the reference system (S) are calculated.

27. A method according to any one of claims 22 to 26, comprising: At least one of the portions of the processed shape image (40) to which the mask device (15) is to be applied is identified based on instructions manually provided by an operator, or based on a shape (150) of the workpiece (50) having a corresponding contour (160), the shape (150) of the workpiece (50) being shown in a processed shape image (30) obtained by processing the image (20) with the aid of an image recognition algorithm.

28. The method according to claim 27, wherein: The image recognition algorithm is a deep learning algorithm configured to process the image (20) and recognize and obtain the processed shape image (30), wherein the processed shape image (30) includes a shape (150) of the workpiece (50) having a corresponding contour (160).

29. The method according to any one of claims 22 to 28, wherein: The workpiece (50, 51, 52) to be processed comprises one of sheet metal (50), sheet metal, semi-finished product (51), sheet metal scrap (52) or sheet metal scrap.

Images