Surface roughness estimation device for workpiece machined surfaces

Abstract

To provide a device for estimating surface roughness of a processed surface of a workpiece which estimates surface roughness of a processed surface of a workpiece at high accuracy.SOLUTION: A device 1 for estimating a surface roughness of a processed surface of a workpiece includes: an image information acquisition unit 13 for acquiring image information for a processed surface formed by applying a machine work to a workpiece; a machining mark information extraction unit 26 for extracting machining mark information related to a machining mark formed on the processed surface of the workpiece, based on the image information acquired by the image information acquisition unit 13; and a surface roughness estimation unit 27 for estimating the surface roughness of the processed surface of the workpiece based on the machining mark information extracted by the machining mark information extraction unit 26.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] The present invention relates to a surface roughness estimation device for a machined surface of a workpiece.

Background Art

[0002] In general, when a workpiece is machined using a tool of a machine tool, there is a risk that the tool will degrade and machining accuracy will decrease. Accordingly, it is extremely critical to maintain the machining accuracy of the workpiece by managing the degradation of the tool and performing correction of the tool as appropriate. As a technique for evaluating the machining accuracy of a workpiece, there is a technique of acquiring an image of the machined surface of the workpiece and estimating the surface roughness of the machined surface of the workpiece that is machined by the tool.

[0003] For example, Patent Document 1 discloses a configuration in which a learned model by a neural network is created in advance on the basis of image information of a machined surface of a workpiece and the surface roughness of the workpiece, appearance image information is acquired from the machined workpiece, and the surface roughness of the machined surface of the workpiece is estimated using the learned model.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of Invention

Problems to be Solved by the Invention

[0005] However, in the configuration disclosed in Patent Document 1, if the conditions for acquiring image information of the workpiece machining surface (e.g., camera placement, ambient light source, and the camera's own shooting error) change, the acquired image information also changes, which may reduce the accuracy of surface roughness estimation. For example, if the incident angle of the illumination light irradiating the workpiece machining surface is large, the proportion of scattered light in the light incident on the camera becomes relatively large, which reduces the brightness contrast in the acquired image information. As a result, the brightness resolution decreases, which may reduce the accuracy of surface roughness estimation.

[0006] This invention has been made in view of the above problems, and aims to provide a workpiece surface roughness estimation device that has excellent accuracy in estimating the surface roughness of the workpiece machined surface. [Means for solving the problem]

[0007] One aspect of the present invention is a surface roughness estimation device for estimating the surface roughness of a machined surface formed by machining a workpiece using a machine tool, An image information acquisition unit that acquires image information of the processed surface, A first image processing unit performs a first image processing on the above image information to clarify the processing marks, In the above image information after the first image processing described above 、 The total area of ​​the processing marks or the total length of the processing marks that are continuous in a predetermined direction. 、 and a measuring unit that measures the number of processing marks present in the above image information, A processing mark information extraction unit calculates at least one of the average area per processing mark or the average length per processing mark in the predetermined direction based on the total area of ​​the processing marks or the total continuous length of the processing marks in a predetermined direction, and the number of processing marks, and extracts it as processing mark information. A surface roughness estimation unit estimates the surface roughness of the processed surface based on a calibration curve that shows the correspondence between the processing mark information and the surface roughness of the processed surface, This is included in the surface roughness estimation device for the machined surface of a workpiece. [Effects of the Invention]

[0008] The inventors of the present invention have obtained a new finding that there is a correlation between machining marks formed on a machined surface when a workpiece is machined by a machine tool and the surface roughness of the machined surface. The present invention is based on this new finding, and in the above-mentioned workpiece machined surface surface roughness estimation device, machining mark information related to the machining marks formed on the machined surface is extracted from the acquired image information of the machined surface, and the surface roughness of the machined surface is estimated based on this machining mark information. Furthermore, the machining mark information related to the machining marks is less affected by changes in the acquisition conditions of the workpiece appearance image compared to the image information of the entire machined surface. Therefore, by estimating the surface roughness of the machined surface based on the above-mentioned machining mark information, the estimation accuracy can be improved.

[0009] As described above, according to the above embodiment, it is possible to provide a workpiece surface roughness estimation device with excellent accuracy in estimating the surface roughness of the workpiece machined surface. [Brief explanation of the drawing]

[0010] [Figure 1] A conceptual diagram showing the configuration of the machine tool in Embodiment 1. [Figure 2] A block diagram of the surface roughness estimation device for the machined surface of a workpiece in Embodiment 1. [Figure 3] Conceptual diagrams showing (a) and (b) the state before luminance distribution conversion in the pre-processing in Embodiment 1, and (c) the state after luminance distribution conversion. [Figure 4] The image of the workpiece machining surface in Embodiment 1 is shown in the following diagrams: (a) an example of the image after grayscale processing, (b) an example of the image after the first image processing, and (c) an example of the image after the second image processing. [Figure 5] A conceptual diagram showing an example of the correspondence between processing mark information and surface roughness in Embodiment 1. [Figure 6] A flowchart for estimating the surface roughness of the workpiece machined surface in Embodiment 1. [Figure 7] A conceptual diagram showing the configuration of the machine tool in Embodiment 2. [Modes for carrying out the invention]

[0011] (Embodiment 1) (1. Machine tool to which the surface roughness estimation device 1 is applied) The surface roughness estimation device 1 for a machined surface of a workpiece in the first embodiment estimates the surface roughness of the machined surface of a workpiece machined by a machine tool. Examples of machine tools for machining workpieces include various grinding devices that perform grinding (e.g., cylindrical grinders, cam grinders, surface grinders, etc.), various cutting devices that perform cutting (e.g., gear machining devices, machining centers, etc.), and polishing devices that perform polishing.

[0012] In the present embodiment, the machine tool is a grinding machine 10 that grinds a workpiece through grinding processes such as rough grinding, finish grinding, fine grinding, and spark-out. Examples of workpieces ground by the grinding machine 10 include simple shaft-shaped members, crankshafts, camshafts, flat plates, and the like. As will be described later, the workpiece of the present embodiment exemplifies a case where a simple shaft-shaped member with a circular cross section is subjected to plunge grinding.

[0013] By machining a workpiece with a tool provided on a machine tool, machining marks are formed on the machined surface of the workpiece. The machining marks are formed along the trajectory of the relative movement of the tool in contact with the workpiece during machining. The specific shape of the machining marks can vary depending on the type of tool, the configuration of the machine tool, the shape of the workpiece, and other factors. When a simple shaft-shaped member with a circular cross section is ground by a cylindrical grinder to be described later, the machining marks are formed in a linear shape extending parallel to the rotational direction of the grinding wheel, and such machining marks are also referred to as grinding streaks.

[0014] As shown in Figure 1, the grinding machine 10 performs grinding on the machined surface W1 of the workpiece W using a grinding wheel 56 as a tool. A control device 12 for controlling the grinding machine 10 is connected to the grinding machine 10. Further, an image information acquisition unit 13 to be described later is attached to the grinding machine 10.

[0015] In the first embodiment, as shown in Figure 1, the grinding machine 10 is a wheel head traverse type cylindrical grinding machine, and is configured to grind the outer peripheral surface, that is, the machined surface W1, of the workpiece W which is a shaft-shaped member. It should be noted that a table traverse type can also be used as the grinding machine 10. In the first embodiment, as shown in Figure 1, the grinding machine 10 mainly includes a bed 51, a headstock 52, a tailstock 53, a traverse base 54, a wheel head 55, a grinding wheel 56, a sizing device 57, and a grinding wheel dressing device 58.

[0016] The bed 51 is fixed on the installation surface of the grinding machine 10. The headstock 52 is provided on the upper surface of the bed 51, on the front side in the X-axis direction (the lower side in Figure 1) and on one end side in the Z-axis direction (the left side in Figure 1). The headstock 52 supports the workpiece W so as to be rotatable around the Z-axis. The workpiece W is rotated by the drive of a motor 52a provided on the headstock 52.

[0017] The tailstock 53 is provided on the upper surface of the bed 51 at a position facing the headstock 52 in the Z-axis direction, that is, on the front side in the X-axis direction (the lower side in Figure 1) and on the other end side in the Z-axis direction (the right side in Figure 1). Thereby, the workpiece W is rotatably supported at both ends by the headstock 52 and the tailstock 53.

[0018] The traverse base 54 is provided on the upper surface of the bed 51 so as to be movable in the Z-axis direction. The traverse base 54 moves by the drive of a motor 54a provided on the bed 51. The wheel head 55 is provided on the upper surface of the traverse base 54 so as to be movable in the X-axis direction. The wheel head 55 moves by the drive of a motor 55a provided on the traverse base 54

[0019] The grinding wheel 56 is formed in a disc shape and is rotatably supported on the grinding wheel base 55. The grinding wheel 56 rotates when driven by a motor 56a provided on the grinding wheel base 55. The grinding wheel 56 is used to grind the outer surface of the workpiece W. If the workpiece W is a cylindrical member, the inner surface of the workpiece W can also be ground. The grinding wheel 56 is constructed by fixing multiple abrasive grains with a binder. The abrasive grains can be general abrasive grains formed from ceramic materials such as alumina or silicon carbide, or superabrasive grains such as diamond or CBN.

[0020] Binders include vitrified (V), resinoid (B), rubber (R), silicate (S), shellac (E), metal (M), electrodeposition (P), and magnesia cement (Mg). Furthermore, the grinding wheel 56 can have a porous structure or a non-porous structure. Depending on the type of binder and the presence or absence of pores, the grinding wheel 56 may have a structure that is elastically deformable or a structure that is almost elastically non-deformable. In the case of an elastically deformable grinding wheel 56, the elastic modulus differs depending on the type of binder, the presence or absence of pores, and the porosity.

[0021] The measuring device 57 measures the dimensions (diameter) of the workpiece W. The measuring device 57 is mounted on the upper surface of the bed 51 so as to be movable in the Z-axis direction. The position of the measuring device 57 in the Z-axis direction is controlled by a feed mechanism 57a provided on the bed 51.

[0022] The grinding wheel correction device 58 corrects the shape of the grinding wheel 56. In other words, the grinding wheel correction device 58 is a device that performs truing and dressing of the grinding wheel 56. Here, truing is a reshaping operation, which is the process of reshaping the grinding wheel 56 to match the shape of the workpiece W when the grinding wheel 56 deteriorates due to grinding, for example, wear, abrasive grain detachment, or abrasive grain crushing, and the process of removing runout of the grinding wheel 56 due to uneven wear. Dressing is a sharpening operation of the grinding wheel 56, which is the process of adjusting the amount of abrasive grain protrusion and creating cutting edges for the abrasive grains. In other words, dressing is the process of correcting dulling, clogging, chipping, etc., and is usually performed after truing.

[0023] The control device 12 includes a CNC device and a PLC device, and as shown in Figure 1, in this embodiment it is connected to the grinding machine 10. The control device 12 controls the motors 52a, 54a, 55a, 56a, etc. in the grinding machine 10 according to an NC program generated based on operation command data such as the shape of the workpiece W, processing conditions, the shape of the grinding wheel 56, and coolant supply timing information. In other words, the control device 12 generates an NC program based on the operation command data when operation command data is input.

[0024] As a result, the control device 12 controls the operation of the grinding machine 10 to grind the machined surface W1 of the workpiece W. Here, the control device 12 grinds the machined surface W1 of the workpiece W until it reaches the finished shape, based on the dimensions (diameter) of the workpiece W measured by the sizing device 57. For this reason, the NC program generated by the control device 12 is generated according to the grinding process, i.e., rough grinding, fine grinding, micro grinding, and sparkout.

[0025] Furthermore, as shown in Figure 1, the control device 12 is capable of sending and receiving signals with the arithmetic unit 20. As a result, the control device 12 controls the motors 52a, 54a, 55a, 56a and the grinding wheel correction device 58, etc., at the timing when it receives a correction signal output from the arithmetic unit 20 to correct the grinding wheel 56, thereby correcting (truing and / or dressing) the grinding wheel 56. Here, the arithmetic unit 20 in this embodiment outputs a correction signal to the control device 12 based on the estimated surface roughness of the machined surface W1, as will be described later.

[0026] (2. Configuration of Surface Roughness Estimation Device 1) Next, the configuration of the surface roughness estimation device 1 will be explained with reference to Figure 1. The surface roughness estimation device 1 consists of an image information acquisition unit 13 and a calculation unit 20.

[0027] As described above, the image information acquisition unit 13 is provided on the grinding machine 10. The image information acquisition unit 13 can be configured with a camera that captures the machined surface W1 of the workpiece W being ground by the grinding machine 10. The position in which the image information acquisition unit 13 is mounted is not limited; in this embodiment 1, as shown in Figure 1, it is mounted on the traverse base 54, but it may be mounted on the grinding wheel base 55 instead. In this embodiment 1, the image information acquisition unit 13 is provided with an illumination device 13a (not shown), and the image information acquisition unit 13 captures the machined surface W1 of the workpiece W when the light from the illumination device 13a is irradiated onto the machined surface W1.

[0028] The image information acquisition unit 13 captures the entire or a predetermined range of the machined surface W1 of the workpiece W for each grinding operation and outputs the obtained image information to the calculation unit 20 described later. The image information acquired by the image information acquisition unit 13 may have multiple color channels or a single color channel.

[0029] The computing unit 20 is composed of a computer and a storage device capable of executing a predetermined program, and as shown in Figure 2, it includes a preprocessing unit 21, a first image processing unit 22, a second image processing unit 23, a first measurement unit 24, a second measurement unit 25, a processing mark information extraction unit 26, a surface roughness estimation unit 27, a measurement result storage unit 28, a processing mark information storage unit 29, a correspondence relationship storage unit 30, and an output unit 31.

[0030] The preprocessing unit 21 performs preprocessing on the image information acquired by the image information acquisition unit 13. In this embodiment 1, if the image information has multiple color channels, the preprocessing unit 21 performs grayscale processing on the image information, discarding the color information and converting it into grayscale image information. An example of the converted image information is shown in Figure 4(a). Alternatively, instead of performing grayscale processing in the preprocessing unit 21, at least one color channel data from the multiple color channels may be used as image information for subsequent processing.

[0031] In this embodiment 1, the preprocessing unit 21 further transforms and normalizes the luminance distribution in the image information so that the average luminance value and the standard deviation of luminance in the image information are within a predetermined range. It is preferable that this transformation is performed so that the luminance of pixels in the image information is distributed as broadly as possible across the luminance range, within a range where at least a portion of the resulting pixel luminance exceeds the luminance range (in this embodiment 1, within a value of 0 to 255), and the image does not become so-called overexposed or underexposed.

[0032] For example, the luminance distribution transformation in the preprocessing unit 21 can be performed as follows. First, as shown in Figure 3(a), the luminance distribution A0 in the image information of the processed surface W1 at the start of processing is obtained. The obtained luminance distribution A0 usually shows a distribution that is roughly close to a normal distribution. Then, the luminance distribution is transformed (normalized) using the luminance transformation formula shown in equation (2), which is obtained by rearranging the normalization formula for the normal distribution shown in equation (1) below, to obtain the transformed luminance distribution As shown in Figure 3(c). Hereinafter, μ is the mean luminance and σ is the standard deviation. (x0-μ0) / σ0=(x'0-μ s ) / σ s (1) x'0 = σ s / σ0*(x0-μ0)+μ s (2) Furthermore, even if the brightness distribution A0 in the image information of the processed surface W1 at the start of processing is as shown in Figure 3(b), by converting (normalizing) the brightness distribution using the above brightness conversion formula, a converted brightness distribution As similar to that shown in Figure 3(c) can be obtained.

[0033] The conversion formula in formula (2) may be one created at the start of machining and used in the subsequent pre-processing unit 21 for the workpiece W, or the conversion formula in formula (2) may be created and used in the pre-processing unit 21 each time image information of the machined surface W1 is acquired for the workpiece W.

[0034] The first image processing unit 22 performs first image processing on the image information after the above preprocessing. The first image processing is a process to clarify the machining marks (grinding marks in this embodiment 1) formed on the machined surface W1. Examples of first image processing to clarify machining marks include edge filtering and Sobel filtering, but in this embodiment 1, Sobel filtering is adopted as the first image processing. Because Sobel filtering can emphasize edges in the image with lower noise than other parts, it is suitable for clarifying machining marks formed on the machined surface W1.

[0035] Sobel filtering includes both horizontal and vertical Sobel filtering. For the first image processing, either one can be used depending on the direction of formation of the processing marks (grinding marks), or both Sobel filtering processes can be used in combination. Furthermore, depending on the type of filter used in the first image processing, the image information after the above preprocessing may be rotated by a predetermined angle before performing the first image processing.

[0036] If the direction of formation of machining marks (grinding marks) is not clear in the pre-processed image information, it is preferable to combine both horizontal and vertical Sobel filtering as the first image processing. In this case, after performing the first image processing by rotating the image information by predetermined angles, subsequent processing can be carried out using the processing result of the first image processing at the image rotation angle that maximizes the length L1 of the machining marks (grinding marks S1) described later.

[0037] In this embodiment 1, as shown in Figure 4(a), the direction in which grinding ridges S1 are formed is the vertical direction (X direction) of the image. Therefore, it is preferable to select a horizontal Sobel filter, which is perpendicular to the formation direction, as the first image processing in the first image processing unit 22. This further emphasizes the edges of the grinding ridges, making them even clearer. An example of the image information after the first image processing in the first image processing unit 22 is shown in Figure 4(b). Note that the first image processing can be performed multiple times. When performing the first image processing multiple times, different filter processing may be included for each step.

[0038] The second image processing unit 23 performs a second image processing on the image information after the first image processing. This second image processing binarizes the image information after the first image processing into black and white. The threshold for binarization in the second image processing is not limited and can be changed as appropriate depending on the image information after the first image processing and the state of the processing marks (grinding marks). An example of the image information after the second image processing in the second image processing unit 23 is shown in Figure 4(c).

[0039] In the image information after the second image processing in the second image processing unit 23, the processing marks (grinding marks S1) can be seen as regions consisting of a series of white pixels. Here, "regions consisting of a series of white pixels" refers to a block of white pixels adjacent to each other in the binarized image, where only white images are connected. For example, in the example of image information after the second image processing (binarization) shown in Figure 4(c), one grinding mark S1 can be seen as a single white line consisting of a series of white pixels in the X direction. In this embodiment 1, in the binarized image shown in Figure 4(c), the difference in size in the Y direction, i.e., the width, of each grinding mark S1 is reduced.

[0040] The first measurement unit 24 measures the total area of ​​white pixels in the image information after the second image processing, or the total length of continuous white pixels in a predetermined direction. The area of ​​white pixels in the image information after the second image processing corresponds to the processing marks (grinding marks S1) on the processed surface W. The total length of continuous white pixels in a predetermined direction refers to the sum of the lengths L1 of all grinding marks S1 in the X direction, as shown in Figure 4(c), that is, the sum of the lengths of all white pixels in the X direction. In this embodiment 1, the grinding machine 10 extends the grinding marks S1, which are processing marks, in the X direction as grinding is repeated, but the change in width in the Y direction is small. Therefore, in this embodiment 1, the total area of ​​the white pixel area strongly reflects the length of the grinding marks S1. The measurement results from the first measurement unit 24 are stored in the measurement result storage unit 28, which will be described later.

[0041] The second measurement unit 25 defines a region consisting of consecutive white pixels in the image information after the second image processing as a single processing mark (grinding ridge S1), and measures the number of grinding ridges S1 present in the image information as the number of processing marks (number of grinding ridges). The measurement results from the second measurement unit 25 are stored in the measurement result storage unit 28, which will be described later.

[0042] The processing mark information extraction unit 26 extracts processing mark information related to processing marks (grinding marks S1) formed on the processed surface W1 based on the image information. This image information may be any of the following: image information acquired by the image information acquisition unit 13, image information after pre-processing by the pre-processing unit 21, image information after first image processing by the first image processing unit 22, or image information after second image processing by the second image processing unit 23.

[0043] The processing mark information extraction unit 26 may preferably extract measurement results measured by the first measurement unit 24 based on the image information as processing mark information. More preferably, the processing mark information extraction unit 26 may extract the average area per processing mark or the average length per processing mark in a predetermined direction in the image information, based on the measurement results of the first measurement unit 24 and the measurement results of the second measurement unit 25, as processing mark information.

[0044] In this embodiment 1, the processing mark information extraction unit 26 extracts the average area of ​​each processing mark (grinding groove S1) in the image information based on the measurement results obtained by the first measurement unit 24, which measures the total area of ​​white pixels in the image information after the second image processing, and the measurement results obtained by the second measurement unit 25. The processing mark information extracted by the processing mark information extraction unit 26 is stored in the processing mark information storage unit 29, which will be described later.

[0045] The surface roughness estimation unit 27 estimates the surface roughness of the machined surface W1 based on the machining mark information extracted by the machining mark information extraction unit 26. Surface roughness estimation can be performed based on a pre-created correspondence between machining mark information and surface roughness. In this embodiment 1, the pre-created correspondence is stored in the correspondence storage unit 30, and the surface roughness is derived and estimated based on the machining mark information extracted by the machining mark information extraction unit 26 using this correspondence. In this embodiment 1, the correspondence storage unit 30 stores a calibration curve P, which shows the correspondence between the average area per grinding mark and the surface roughness, as shown in Figure 5.

[0046] The measurement result storage unit 28, the processing mark information storage unit 29, and the correspondence relationship storage unit 30 consist of rewritable non-volatile memory, and various types of information are stored therein as described above. The output unit 31 outputs the estimation result of the surface roughness estimation unit 27. The output mode of the output unit 31 is not limited, and the estimation result may be displayed on a predetermined display. The output unit 31 may also transmit a correction signal to the control device 12 of the grinding machine 10 to instruct the grinding wheel 56 to perform dressing or truing based on the estimation result.

[0047] (3. Surface roughness estimation flow of surface roughness estimation device 1) Next, the flow for estimating the surface roughness of the machined surface W1 in the surface roughness estimation device 1 of this embodiment 1 will be explained using Figure 6. First, in step S1 shown in Figure 6, the machining of the workpiece W by the grinding machine 10 is started.

[0048] Next, in step S2 shown in Figure 6, the image information acquisition unit 13 acquires image information of the processed surface W1. In this embodiment 1, the image information has a color channel. Then, in step S3, the preprocessing unit 21 converts the image information to grayscale and discards the color information. This creates the image shown in Figure 4(a).

[0049] Next, in step S4, the preprocessing unit 21 performs a luminance distribution modification process. In this embodiment 1, the average luminance μ0 and the standard luminance deviation σ0 are calculated from the image information of the processed surface W1 acquired at the start of processing to create the above-mentioned conversion formula (2), and the luminance distribution of the processed surface W1 is modified and normalized using the conversion formula (2). As a result, the average luminance μ in the image information is changed. s and the luminance standard deviation σ s By modifying each of these parameters and normalizing the luminance distribution, overexposure and underexposure are prevented in the modified image information.

[0050] Subsequently, in step S5 shown in Figure 6, the first image processing unit 22 performs a Sobel filter as the first image processing step. As a result, the edges of the grinding marks S1 are emphasized, as shown in Figure 4(b).

[0051] Then, in step S6 shown in Figure 6, the second image processing unit 23 performs a black and white binarization process as a second image processing step based on a threshold that has been set appropriately. As a result, in the image information after binarization, the grinding marks S1 are shown as white areas, and the other parts are shown as black areas, as shown in Figure 4(c).

[0052] Next, in step S7 shown in Figure 6, the first measurement unit 24 measures the total area of ​​the grinding grooves S1 in the image information and stores the measurement result in the measurement result storage unit 28. Then, in step S8, the second measurement unit 25 measures the number of grinding grooves S1 in the image information and stores the measurement result in the measurement result storage unit 28.

[0053] Subsequently, in step S9 shown in Figure 6, the machining mark information extraction unit 26 calculates the average area per grinding groove as machining mark information and stores the calculation result in the machining mark information storage unit 29. Then, in step S10, the surface roughness estimation unit 27 estimates the surface roughness of the machined surface W1 using the calibration curve P pre-stored in the correspondence relationship storage unit 30 shown in Figure 5, based on the average area per grinding groove as machining mark information stored in the machining mark information storage unit 29. Finally, the output unit 31 outputs the estimation result and terminates the control flow.

[0054] (4. Effects and Benefits) The effects and advantages of the surface roughness estimation device 1 of claim 1 of this application are described below. According to the surface roughness estimation device 1 of claim 1 of this application, processing mark information related to grinding marks S1, which are processing marks formed on the processing surface W1, is extracted from the acquired image information of the processing surface W1, and the surface roughness of the processing surface is estimated based on said processing mark information. Furthermore, the processing mark information related to processing marks is less affected by changes in the acquisition conditions of the external image of the workpiece W compared to the image information of the entire processing surface W1. Therefore, by estimating the surface roughness of the processing surface W1 based on said processing mark information, the estimation accuracy can be improved.

[0055] Furthermore, in this embodiment 1, the system includes a first image processing unit 22 that performs a first image processing to clarify the grinding ridges S1, which are processing marks, in the image information acquired by the image information acquisition unit 13. The processing mark information extraction unit 26 then extracts processing mark information based on the image information after the first image processing. As a result, the processing mark information is extracted with the grinding ridges S1 clearly defined, thereby improving the detection accuracy of the grinding ridges S1.

[0056] Furthermore, in this embodiment 1, the first image processing in the first image processing unit 22 includes Sobel filtering. This further enhances the edges of the grinding marks S1, making the grinding marks S1 clearer and further improving the detection accuracy of the grinding marks S1.

[0057] Furthermore, in this embodiment 1, the system further includes a second image processing unit 23 that binarizes the image information after the first image processing into black and white, a first measurement unit 24 that measures the total area of ​​the white pixel region in the image information after the second image processing or the total length L1 of the continuous white pixels in a predetermined direction, and a second measurement unit that defines a region of continuous white pixels in the image information after the second image processing as one grinding streak S1 and measures the number of grinding streaks S1 present in the image information as the number of processing marks.The processing mark information extraction unit 26 extracts the average area per grinding streak or the average length in the X direction per grinding streak in the image information as processing mark information, based on the measurement results of the first measurement unit 24 and the measurement results of the second measurement unit 25.As a result, the processing mark information includes information on processing marks (grinding streaks S1) that have been detected with high accuracy, thereby further improving the accuracy of surface roughness estimation.

[0058] Furthermore, in this embodiment 1, a preprocessing unit 21 is provided that performs a preprocessing operation to normalize the luminance distribution in the image information so that the average luminance value and standard deviation of luminance in the image information are within a predetermined range, as a preprocessing operation for extracting processing mark information. This makes it possible to suppress the influence of changes in the conditions under which image information is acquired by the image information acquisition unit 13, such as changes in the illumination light illuminating the processed surface W1, on the extraction of processing mark information, thereby further improving the accuracy of surface roughness estimation. In addition, by performing the above preprocessing operation, it is possible to mix image information acquired when the processed surface W1 of the workpiece W is dry and image information acquired when the processed surface W1 is wet, making it easier to handle the image information. Furthermore, by performing the above preprocessing operation, it is also possible to quantify the processing marks.

[0059] In this embodiment 1, the processing mark information extraction unit 26 extracts the average area per grinding groove as processing mark information using the measurement results of the first measurement unit 24 and the measurement results of the second measurement unit 25. However, instead, the processing mark information extraction unit 26 may extract the total area of ​​the grinding groove S1 or the total length of the grinding groove S1, which are the measurement results of the first measurement unit 24, as processing mark information, without having the second measurement unit 25. In this case, the accuracy of surface roughness estimation may be slightly lower compared to this embodiment 1, which uses the measurement results of the second measurement unit 25. However, since calculations by the second measurement unit 25 are unnecessary compared to this embodiment 1, the computational load can be reduced and the estimation process can be sped up. In contrast, in this embodiment 1, the accuracy of surface roughness estimation is further improved by using the measurement results of the second measurement unit 25.

[0060] As described above, according to the above embodiment, it is possible to provide a surface roughness estimation device 1 for a workpiece machined surface W1 that has excellent accuracy in estimating the surface roughness of the workpiece machined surface.

[0061] (Embodiment 2) In Embodiment 1 described above, a grinding machine 10 was used as a machine tool, but in Embodiment 2, a cutting machine 90 shown in Figure 7 is used instead. In Embodiment 2, the cutting machine 90 is a lathe that performs cutting on the machined surface W1 of a workpiece W using a cutting tool 96. A control device 12 that controls the cutting machine 90 is connected to the cutting machine 90, as in Embodiment 1. In Embodiment 2, the cutting machine 90 rotates the workpiece W under the control of the control device 12 and turns the workpiece W by bringing the cutting tool 96 into contact with the workpiece W and moving it relative to it. As a result, cutting marks made by the cutting tool 96 are formed on the machined surface W2 of the workpiece W as machining marks. In Embodiment 2, equivalent components from Embodiment 1 are denoted by the same reference numerals and their descriptions are omitted.

[0062] The cutting machine 90 includes, for example, a bed 51, a headstock 52, a tailstock 53, and a traverse base 54. The tool rest 95 is mounted on the upper surface of the traverse base 54 so as to be movable in the X-axis direction. The tool rest 95 is moved by the drive of a motor 55a provided on the traverse base 54. A cutting tool 96 is supported on the tool rest 95. Furthermore, the cutting machine 90 includes a sizing device 57.

[0063] In this second embodiment as well, the image information acquisition unit 13 is provided on the cutting machine 90 and captures an image of the machined surface W2 of the workpiece W being cut by the cutting machine 90 while the light from the illumination device 13a is irradiated onto it. The image is then processed by a computing device 20 having the same configuration as in the first embodiment, and the surface roughness of the machined surface W2 of the workpiece W is estimated.

[0064] In Embodiment 2, the average area per machining mark in Embodiment 1 can be replaced with the average pitch between machining marks, and the average length per machining mark in Embodiment 1 can be replaced with the average feed rate of the bit 96 per machining mark. Furthermore, Embodiment 2 achieves the same effects and advantages as Embodiment 1.

[0065] The present invention is not limited to the embodiments described above, and can be applied to various embodiments without departing from its spirit.

[0066] In Embodiment 1, the machine tool was a grinding machine 10, and in Embodiment 2, the machine tool was a cutting machine 90 which is a lathe. However, the machine tool is not limited to these and can be any other machine tool, for example, a cutting machine which is a milling machine. [Explanation of Symbols]

[0067] 1. Surface roughness estimation device 10 Grinding machine 12 Control device 13 Image Information Acquisition Unit 20 Arithmetic unit 21 Pre-processing section 22 First Image Processing Unit 23. Second Image Processing Unit 24. First Measurement Unit 25. 2nd BC Section 26 Machining mark information extraction part 27 Surface roughness estimation section 57 Sizing device 58 Grinding wheel correction device 90 cutting machine

Claims

1. A surface roughness estimation device for estimating the surface roughness of a machined surface formed by machining a workpiece using a machine tool, An image information acquisition unit that acquires image information of the processed surface, A first image processing unit performs a first image processing on the above image information to clarify the processing marks, A measuring unit measures the total area of ​​the processing marks or the total length of the processing marks that are continuous in a predetermined direction, and the number of processing marks present in the image information after the first image processing described above. A processing mark information extraction unit calculates at least one of the average area per processing mark or the average length per processing mark in the predetermined direction based on the total area of ​​the processing marks or the total continuous length of the processing marks in a predetermined direction, and the number of processing marks, and extracts it as processing mark information. A surface roughness estimation unit estimates the surface roughness of the processed surface based on a calibration curve that shows the correspondence between the processing mark information and the surface roughness of the processed surface, A device for estimating the surface roughness of a workpiece's machined surface, including the device itself.

2. The above-mentioned first image processing includes Sobel filtering, wherein the workpiece surface roughness estimation apparatus is as described in claim 1.

3. The system further includes a second image processing unit that binarizes the image information after the first image processing into black and white. The surface roughness estimation device for a workpiece machined surface according to claim 1 or 2, wherein the measurement unit includes a first measurement unit that measures the total area of ​​white pixel regions in the image information after the second image processing, or the total length of continuous white pixels in a predetermined direction, as the total area of ​​machined marks in the image information or the total length of continuous white pixels in a predetermined direction.

4. The surface roughness estimation device for a workpiece machined surface according to claim 3, wherein the measurement unit includes a second measurement unit that defines a region consisting of consecutive white pixels in the image information after the second image processing as one of the processing marks, and measures the number of processing marks present in the image information as the number of processing marks.

5. The surface roughness estimation device for a workpiece machined surface according to claim 1 or 2, comprising a preprocessing unit that transforms and normalizes the luminance distribution in the image information so that the average luminance value and standard deviation of luminance in the image information are within a predetermined range, as a preprocessing step for extracting the above-mentioned machining mark information.

6. The surface roughness estimation device for a workpiece machined surface according to claim 1 or 2, wherein the machine tool is a grinding machine or a cutting machine.

Images