Working surface roughness detection method and system based on machine vision

Abstract

The invention provides a working surface roughness detection method and system based on machine vision, and the method comprises the steps: analyzing a first global image of a working surface under the irradiation of reference illumination light, obtaining image shadow distribution feature information, and determining an abnormal illumination sub-region in an illumination space formed by the reference illumination light; on the basis of the abnormal lighting sub-region, the irradiation state of additional lighting light on the working surface is adjusted, and global uniform lighting on the working surface is achieved; analyzing a second global image of the working surface under the common irradiation of the reference illumination light and the additional illumination light to obtain textural feature information and scattered light feature information of the working surface, and characterizing the morphology of the working surface on an image vision and light scattering level; based on the texture feature information and the scattered light feature information, different morphological distribution attribute information of the working surface is determined respectively, so that a roughness plane distribution model of the working surface is obtained through integration, and the accuracy of working surface roughness detection is improved.

Description

Technical Field

[0001] The present invention relates to the field of visual recognition, and in particular to a method and system for detecting the roughness of a working surface based on machine vision. Background Art

[0002] As components of precision instruments, precision parts have high requirements for surface roughness. If the surface roughness of precision parts is large, two adjacent precision parts will form a large friction force during the working process, causing wear problems on precision parts and affecting the working life of precision parts. For this reason, it is necessary to perform roughness detection on the working surface of parts during the production process of precision parts to obtain the global roughness distribution of the working surface, which provides a basis for the deep processing of precision parts. The existing detection of the working surface roughness of precision parts is achieved by a single visual recognition method, that is, the working surface is imaged and analyzed. In this way, the detection accuracy of the working surface roughness directly depends on the resolution of the image captured. The existing image shooting resolution cannot meet the high-precision requirements for the working surface roughness, and it is also impossible to perform multi-dimensional detection and analysis of the roughness of the working surface, which reduces the accuracy and reliability of the working surface roughness detection. Summary of the invention

[0003] The purpose of the present invention is to provide a working surface roughness detection method and system based on machine vision, which analyzes the first global image of the working surface under the illumination of reference illumination light to obtain the image shadow distribution feature information, thereby determining the abnormal illumination sub-area inside the illumination space formed by the reference illumination light, and accurately positioning the area of ​​the working surface that is not fully and effectively illuminated; based on the abnormal illumination sub-area, adjust the illumination state of the additional illumination light on the working surface to achieve global uniform illumination of the working surface; analyze the second global image of the working surface under the common illumination of the reference illumination light and the additional illumination light to obtain the texture feature information and scattered light feature information of the working surface, and characterize the morphology of the working surface at the image vision and light scattering levels; based on the texture feature information and the scattered light feature information, respectively determine the different morphology distribution attribute information of the working surface, thereby integrating to obtain the roughness plane distribution model of the working surface, improve the accuracy of the roughness detection of the working surface, and provide a full range of basis for the refined processing of the working surface.

[0004] The present invention is achieved through the following technical solutions:

[0005] A method for detecting the roughness of a working surface based on machine vision, comprising:

[0006] Acquire a first global image of the work surface under the illumination of the reference illumination light, analyze the first global image, and obtain image shadow distribution characteristic information of the first global image; determine an abnormal illumination sub-region within the illumination space formed by the reference illumination light based on the image shadow distribution characteristic information;

[0007] Based on the attribute information of the abnormal illumination sub-region, adjusting the illumination state of the additional illumination light on the working surface; acquiring a second global image of the working surface under the illumination of the reference illumination light and the additional illumination light; analyzing the second global image to obtain texture feature information and scattered light feature information of the working surface;

[0008] Based on the texture feature information, first morphology distribution attribute information of the working surface is determined; based on the scattered light feature information, second morphology distribution attribute information of the working surface is determined; based on the first morphology distribution attribute information and the second morphology distribution attribute information, a roughness plane distribution model of the working surface is generated.

[0009] Optionally, acquiring a first global image of the work surface under the illumination of the reference illumination light, analyzing the first global image to obtain image shadow distribution feature information of the first global image; and determining an abnormal illumination sub-region inside the illumination space formed by the reference illumination light based on the image shadow distribution feature information, comprises:

[0010] Scanning and photographing a work surface under the illumination of a reference illumination light to obtain a first global image of the work surface; performing image brightness recognition on the first global image to obtain image shadow distribution feature information of the first global image;

[0011] Based on the picture shadow distribution characteristic information, determine the picture shadow area distribution range position of the first global image; map the picture shadow area distribution range position to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area inside the illumination space.

[0012] Optionally, based on the attribute information of the abnormal illumination sub-region, adjusting the illumination state of the additional illumination light on the working surface; acquiring a second global image of the working surface under the common illumination of the reference illumination light and the additional illumination light; and analyzing the second global image to obtain texture feature information and scattered light feature information of the working surface, including:

[0013] Based on the distribution position information and light intensity information of the abnormal illumination sub-area, adjusting the irradiation position and irradiation intensity of the additional illumination light irradiated on the work surface;

[0014] Performing binocular photography on the working surface under the illumination of the reference illumination light and the additional illumination light to obtain a global binocular image of the working surface as the second global image; and generating a corresponding global three-dimensional image based on the binocular parallax of the global binocular image;

[0015] Perform pixel texture recognition on the global three-dimensional image to obtain texture depth and texture direction feature information of the working surface; perform background light distribution extraction processing on the global three-dimensional image to obtain a background light distribution image of the global three-dimensional image; perform light intensity distribution uniformity recognition on the background light distribution image to obtain scattered light distribution feature information of the working surface.

[0016] Optionally, adjusting the irradiation position and irradiation intensity of the additional illuminating light irradiated on the work surface based on the distribution position information and the light intensity information of the abnormal illumination sub-area includes:

[0017] Step S1, using the following formula (1), according to the distribution position information and light intensity information of the abnormal illumination sub-area, determine the irradiation position of the additional illumination light on the working surface,

[0018]

[0019] In the above formula (1), (X 0 ,Y 0 ) represents the coordinate point of the irradiation position of the working surface with the additional illumination light; Indicates coordinate points Coordinate point and coordinate points The coordinate point of the hammer point with the line of the connecting line as a perpendicular line; [X(k), Y(k)] represents the kth abnormal lighting sub-area position coordinate point in the distribution position information of the abnormal lighting sub-area; n represents the total number of abnormal lighting sub-area position coordinate points in the distribution position information of the abnormal lighting sub-area; It means to put the value of k from 1 to n into the brackets to get the maximum value in the brackets; It means to put the value of k from 1 to n into the brackets to get the minimum value in the brackets;

[0020] Step S2, using the following formula (2), according to the distribution position information and light intensity information of the abnormal illumination sub-area, determine the illumination intensity of the additional illumination light irradiated on the work surface,

[0021]

[0022] In the above formula (2), E represents the intensity of the additional illumination light irradiated on the working surface; E qrepresents the light intensity of the abnormal illumination sub-region; L{[X(k),Y(k)],(X 0 ,Y 0 )} means to get the coordinate point [X(k), Y(k)] and the coordinate point (X 0 ,Y 0 ) between the two; E 0 Indicates the minimum light intensity of a circle with a unit distance of irradiation diameter; D 0 Represents unit distance;

[0023] Step S3, using the following formula (3), according to the intensity of the additional illumination light irradiated on the work surface and the duration of the additional illumination light, whether it is necessary to switch from the currently used additional illumination light to other additional illumination lights,

[0024]

[0025] In the above formula (3), Z 2 Indicates the judgment value for switching from the currently used additional lighting to other additional lighting; T 0 It indicates the duration of continuous irradiation at the minimum light intensity of a circle with a diameter of unit distance; t indicates the current time; t 0 represents the time when the additional illumination light starts to irradiate; T h Indicates the switching time of the additional lighting;

[0026] If Z 2 =1, it is necessary to switch from the currently used additional lighting to other additional lighting;

[0027] If Z 2 =0, there is no need to switch from the currently used additional lighting light to other additional lighting lights.

[0028] Optionally, based on the texture feature information, determining first topography distribution attribute information of the working surface; based on the scattered light feature information, determining second topography distribution attribute information of the working surface; based on the first topography distribution attribute information and the second topography distribution attribute information, generating a roughness plane distribution model of the working surface, comprising:

[0029] Determine the depth distribution attribute information of the topographic profile undulation of the working surface based on the texture depth and texture direction characteristic information of the working surface; determine the interval distribution attribute information of the topographic profile undulation of the working surface based on the scattered light distribution characteristic information of the working surface;

[0030] The depth distribution attribute information of the topography profile fluctuation and the interval distribution attribute information of the topography profile fluctuation are integrated into the plane space corresponding to the working surface, so as to generate a roughness plane distribution model of the working surface.

[0031] A working surface roughness detection system based on machine vision, comprising:

[0032] A first visual recognition module is used to obtain a first global image of the work surface under the illumination of a reference illumination light, analyze the first global image, and obtain image shadow distribution feature information of the first global image;

[0033] An abnormal lighting area identification module, used to determine the abnormal lighting sub-area inside the lighting space formed by the reference lighting light based on the image shadow distribution feature information;

[0034] an illumination light adjustment module, configured to adjust the illumination state of the additional illumination light on the work surface based on the attribute information of the abnormal illumination sub-region;

[0035] A second visual recognition module is used to obtain a second global image of the working surface under the combined illumination of the reference illumination light and the additional illumination light; and analyze the second global image to obtain texture feature information and scattered light feature information of the working surface;

[0036] A surface morphology analysis module, configured to determine first morphology distribution attribute information of the working surface based on the texture feature information; and to determine second morphology distribution attribute information of the working surface based on the scattered light feature information;

[0037] A surface roughness distribution characterization module is used to generate a roughness plane distribution model of the working surface based on the first morphology distribution attribute information and the second morphology distribution attribute information.

[0038] Optionally, the first visual recognition module is used to obtain a first global image of the work surface under the illumination of the reference illumination light, and analyze the first global image to obtain image shadow distribution feature information of the first global image, including:

[0039] Scanning and photographing a work surface under the illumination of a reference illumination light to obtain a first global image of the work surface; performing image brightness recognition on the first global image to obtain image shadow distribution feature information of the first global image;

[0040] The abnormal lighting area identification module is used to determine the abnormal lighting sub-area inside the lighting space formed by the reference lighting light based on the picture shadow distribution feature information, including:

[0041] Based on the picture shadow distribution characteristic information, determine the picture shadow area distribution range position of the first global image; map the picture shadow area distribution range position to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area inside the illumination space.

[0042] Optionally, the illumination light adjustment module is used to adjust the illumination state of the additional illumination light on the work surface based on the attribute information of the abnormal illumination sub-region, including:

[0043] Based on the distribution position information and light intensity information of the abnormal illumination sub-area, adjusting the irradiation position and irradiation intensity of the additional illumination light irradiated on the work surface;

[0044] The second visual recognition module is used to obtain a second global image of the working surface under the combined illumination of the reference illumination light and the additional illumination light; and analyze the second global image to obtain texture feature information and scattered light feature information of the working surface, including:

[0045] Performing binocular photography on the working surface under the illumination of the reference illumination light and the additional illumination light to obtain a global binocular image of the working surface as the second global image; and generating a corresponding global three-dimensional image based on the binocular parallax of the global binocular image;

[0046] Perform pixel texture recognition on the global three-dimensional image to obtain texture depth and texture direction feature information of the working surface; perform background light distribution extraction processing on the global three-dimensional image to obtain a background light distribution image of the global three-dimensional image; perform light intensity distribution uniformity recognition on the background light distribution image to obtain scattered light distribution feature information of the working surface.

[0047] Optionally, the surface morphology analysis module is used to determine first morphology distribution attribute information of the working surface based on the texture feature information; and to determine second morphology distribution attribute information of the working surface based on the scattered light feature information, including:

[0048] Determine the depth distribution attribute information of the topographic profile undulation of the working surface based on the texture depth and texture direction characteristic information of the working surface; determine the interval distribution attribute information of the topographic profile undulation of the working surface based on the scattered light distribution characteristic information of the working surface;

[0049] The surface roughness distribution characterization module is used to generate a roughness plane distribution model of the working surface based on the first morphology distribution attribute information and the second morphology distribution attribute information, including:

[0050] The depth distribution attribute information of the topography profile fluctuation and the interval distribution attribute information of the topography profile fluctuation are integrated into the plane space corresponding to the working surface, so as to generate a roughness plane distribution model of the working surface.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] The machine vision-based working surface roughness detection method and system provided in the present application analyze the first global image of the working surface under the illumination of the reference illumination light to obtain the image shadow distribution feature information, thereby determining the abnormal illumination sub-area within the illumination space formed by the reference illumination light, and accurately locating the area of ​​the working surface that is not fully and effectively illuminated; based on the abnormal illumination sub-area, adjust the illumination state of the additional illumination light on the working surface to achieve global uniform illumination of the working surface; analyze the second global image of the working surface under the common illumination of the reference illumination light and the additional illumination light to obtain the texture feature information and scattered light feature information of the working surface, and characterize the morphology of the working surface at the image vision and light scattering levels; based on the texture feature information and the scattered light feature information, respectively determine the different morphology distribution attribute information of the working surface, thereby integrating the roughness plane distribution model of the working surface, improving the accuracy of the roughness detection of the working surface, and providing a comprehensive basis for the refined processing of the working surface. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the prior art descriptions. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0054] Figure 1 A schematic flow chart of a method for detecting working surface roughness based on machine vision provided by the present invention.

[0055] Figure 2 A schematic structural diagram of a working surface roughness detection system based on machine vision provided by the present invention. DETAILED DESCRIPTION

[0056] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below in conjunction with the accompanying drawings. It is to be understood that the specific embodiments described herein are only used to explain the present application, rather than to limit the present application. It should also be noted that, for ease of description, only some structures related to the present application are shown in the accompanying drawings, rather than all structures. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0057] The terms "including" and "having" and any variations thereof in this application are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device comprising a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products or devices.

[0058] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0059] See also Figure 1 As shown, an embodiment of the present application provides a method for detecting the roughness of a working surface based on machine vision. The method for detecting the roughness of a working surface based on machine vision includes:

[0060] Acquire a first global image of the work surface under the illumination of the reference illumination light, analyze the first global image, and obtain image shadow distribution characteristic information of the first global image; determine an abnormal illumination sub-area within the illumination space formed by the reference illumination light based on the image shadow distribution characteristic information;

[0061] Based on the attribute information of the abnormal illumination sub-region, adjusting the illumination state of the additional illumination light on the work surface; obtaining a second global image of the work surface under the common illumination of the reference illumination light and the additional illumination light; analyzing the second global image to obtain texture feature information and scattered light feature information of the work surface;

[0062] Based on the texture feature information, first morphology distribution attribute information of the working surface is determined; based on the scattered light feature information, second morphology distribution attribute information of the working surface is determined; based on the first morphology distribution attribute information and the second morphology distribution attribute information, a roughness plane distribution model of the working surface is generated.

[0063] The beneficial effects of the above embodiments are as follows: the working surface roughness detection method based on machine vision analyzes the first global image of the working surface under the illumination of the reference illumination light to obtain the image shadow distribution feature information, thereby determining the abnormal illumination sub-area within the illumination space formed by the reference illumination light, and accurately locating the area of ​​the working surface that is not fully and effectively illuminated; based on the abnormal illumination sub-area, adjusting the illumination state of the additional illumination light on the working surface to achieve global uniform illumination of the working surface; analyzing the second global image of the working surface under the common illumination of the reference illumination light and the additional illumination light to obtain the texture feature information and scattered light feature information of the working surface, and characterizing the morphology of the working surface at the image vision and light scattering levels; based on the texture feature information and the scattered light feature information, respectively determining the different morphology distribution attribute information of the working surface, thereby integrating the roughness plane distribution model of the working surface, improving the accuracy of the roughness detection of the working surface, and providing a full range of basis for the refined processing of the working surface.

[0064] In another embodiment, a first global image of a work surface under the illumination of a reference illumination light is acquired, and the first global image is analyzed to obtain image shadow distribution characteristic information of the first global image; and based on the image shadow distribution characteristic information, an abnormal illumination sub-region within the illumination space formed by the reference illumination light is determined, including:

[0065] Scanning and photographing a working surface under the illumination of a reference illumination light to obtain a first global image of the working surface; performing image brightness recognition on the first global image to obtain image shadow distribution feature information of the first global image;

[0066] Based on the image shadow distribution characteristic information, the image shadow area distribution range position of the first global image is determined; the image shadow area distribution range position is mapped to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area within the illumination space.

[0067] The beneficial effect of the above embodiment is that when the working surface of the precision parts is unevenly illuminated during the visual shooting process, the image of the working surface obtained will have obvious dark areas, so that the image of the working surface corresponding to the obvious dark areas cannot be identified and analyzed. For this reason, the working surface under the irradiation of the reference illumination light is scanned and photographed to obtain a first global image of the working surface; wherein the reference illumination light may be, but is not limited to, a collimated illumination light with a uniform cross-sectional light intensity distribution, so that the first global image can fully reflect the light intensity distribution of the working surface under the irradiation of the reference illumination light for characterization. The first global image is analyzed for the screen brightness value to obtain the screen shadow distribution feature information of the first global image; if the average brightness of a certain screen area of ​​the first global image is less than the preset brightness threshold, the corresponding screen area is determined as the screen shadow area, so that the distribution of all screen shadow areas on the first global image is fully identified and calibrated. In addition, based on the boundary feature information of the shadow distribution of the picture, the distribution range position of the shadow area of ​​the first global image is determined, and the distribution range position of the shadow area is mapped to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area inside the illumination space, so that the abnormal illumination sub-area corresponds one-to-one with the shadow area of ​​the picture, and subsequently the shadow area of ​​the picture can be directionally eliminated by performing fill light illumination in the abnormal illumination sub-area.

[0068] In another embodiment, based on the attribute information of the abnormal illumination sub-region, adjusting the illumination state of the additional illumination light on the work surface; acquiring a second global image of the work surface under the common illumination of the reference illumination light and the additional illumination light; and analyzing the second global image to obtain texture feature information and scattered light feature information of the work surface, including:

[0069] Based on the distribution position information and light intensity information of the abnormal illumination sub-area, adjusting the irradiation position and irradiation intensity of the additional illumination light irradiated on the work surface;

[0070] Performing binocular photography on the working surface under the illumination of the reference illumination light and the additional illumination light to obtain a global binocular image of the working surface as the second global image; and generating a corresponding global three-dimensional image based on the binocular parallax of the global binocular image;

[0071] Pixel texture recognition is performed on the global three-dimensional image to obtain texture depth and texture direction feature information of the working surface; background light distribution extraction processing is performed on the global three-dimensional image to obtain a background light distribution image of the global three-dimensional image; light intensity distribution uniformity recognition is performed on the background light distribution image to obtain scattered light distribution feature information of the working surface.

[0072] The beneficial effect of the above embodiment is that the corresponding area of ​​the work surface is supplemented with light based on the distribution position information and light intensity information of the abnormal illumination sub-area, that is, the illumination position and light intensity of the additional illumination light are illuminated on the work surface, so that the shadow elimination illumination of the work surface can be accurately directed to ensure that the work surface is uniformly illuminated as a whole. In addition, the work surface under the common illumination of the reference illumination light and the additional illumination light is photographed and converted by binoculars to obtain a global three-dimensional image of the work surface, thereby performing a three-dimensional characterization of the morphological contour of the work surface. Then, the global three-dimensional image is subjected to pixel texture recognition, background light distribution extraction processing, and light intensity distribution uniformity recognition to obtain the texture depth and texture direction feature information of the work surface and the scattered light distribution feature information, and the surface undulation and roughness state of the work surface is characterized from the two aspects of image pixel features and light scattering features.

[0073] In another embodiment, adjusting the irradiation position and irradiation intensity of the additional illuminating light irradiated on the work surface based on the distribution position information and the light intensity information of the abnormal illumination sub-region includes:

[0074] Step S1, using the following formula (1), according to the distribution position information and light intensity information of the abnormal illumination sub-area, determine the irradiation position of the work surface for irradiating the additional illumination light,

[0075]

[0076] In the above formula (1), (X 0 ,Y 0 ) represents the coordinate point of the irradiation position of the work surface with the additional illumination light; Indicates coordinate points Coordinate point and coordinate points The coordinate point of the hammer point with the line of the connecting line as the perpendicular line; [X(k), Y(k)] represents the kth abnormal lighting sub-region position coordinate point in the distribution position information of the abnormal lighting sub-region; n represents the total number of abnormal lighting sub-region position coordinate points in the distribution position information of the abnormal lighting sub-region; It means to put the value of k from 1 to n into the brackets to get the maximum value in the brackets; It means to put the value of k from 1 to n into the brackets to get the minimum value in the brackets;

[0077] Step S2, using the following formula (2), according to the distribution position information and light intensity information of the abnormal illumination sub-area, determine the illumination intensity of the additional illumination light irradiated on the work surface,

[0078]

[0079] In the above formula (2), E represents the intensity of the additional illumination light irradiated on the working surface; E q represents the light intensity of the abnormal illumination sub-region; L{[X(k),Y(k)],(X 0 ,Y 0 )} means to get the coordinate point [X(k), Y(k)] and the coordinate point (X 0 ,Y 0 ) between the two; E 0 Indicates the minimum light intensity of a circle with a unit distance of irradiation diameter; D 0 Represents unit distance;

[0080] Step S3, using the following formula (3), according to the illumination intensity and continuous illumination time of the additional illumination light irradiating the work surface, whether it is necessary to switch from the currently used additional illumination light to other additional illumination lights,

[0081]

[0082] In the above formula (3), Z 2 Indicates the judgment value for switching from the currently used additional lighting to other additional lighting; T 0 It indicates the duration of continuous irradiation at the minimum light intensity of a circle with a diameter of unit distance; t indicates the current time; t 0 Indicates the time when the additional lighting light starts to irradiate; T h Indicates the switching time of the additional lighting;

[0083] If Z 2 =1, it is necessary to switch from the currently used additional lighting to other additional lighting;

[0084] If Z 2 =0, there is no need to switch from the currently used additional lighting light to other additional lighting lights.

[0085] The beneficial effect of the above embodiment is that, using the above formula (1), according to the distribution position information and light intensity information of the abnormal lighting sub-area, the irradiation position of the additional lighting light irradiating the working surface is determined, thereby ensuring that the irradiation position of the additional lighting light covers the abnormal lighting sub-area as much as possible; then using the above formula (2), according to the distribution position information and light intensity information of the abnormal lighting sub-area, the irradiation intensity of the additional lighting light irradiating the working surface is determined, thereby ensuring that the light intensity can improve the lighting condition of the abnormal lighting sub-area; then using the above formula (3), according to the irradiation intensity and continuous irradiation time of the additional lighting light irradiating the working surface, whether it is necessary to switch from the currently used additional lighting light to other additional lighting lights, thereby regularly switching the additional lighting light, ensuring the reliable operation of the additional lighting equipment and extending the service life of the additional lighting equipment.

[0086] In another embodiment, based on the texture feature information, first topography distribution attribute information of the working surface is determined; based on the scattered light feature information, second topography distribution attribute information of the working surface is determined; based on the first topography distribution attribute information and the second topography distribution attribute information, a roughness plane distribution model of the working surface is generated, including:

[0087] Determine the depth distribution attribute information of the topographic profile undulation of the working surface based on the texture depth and texture direction characteristic information of the working surface; determine the interval distribution attribute information of the topographic profile undulation of the working surface based on the scattered light distribution characteristic information of the working surface;

[0088] The depth distribution attribute information of the topographic profile fluctuation and the interval distribution attribute information of the topographic profile fluctuation are integrated into the plane space corresponding to the working surface, so as to generate a roughness plane distribution model of the working surface.

[0089] The beneficial effect of the above-mentioned embodiment is that the morphological fluctuation of the working surface is the direct cause of the roughness of the working surface, and the morphological fluctuation of the working surface is reflected as texture depth and texture direction at the image level. At the same time, when the light is irradiated to the working surface, the corresponding rough morphology will scatter the light (the scattering characteristics of the light are related to the morphological contour fluctuation interval of the working surface), so that the background light of the image of the working surface will present corresponding light scattering characteristics. For this reason, the texture depth and texture direction feature information of the working surface are used to determine the morphological contour fluctuation depth distribution attribute information of the working surface; based on the scattered light distribution feature information of the working surface, the morphological contour fluctuation interval distribution attribute information of the working surface is determined, so that the surface undulation roughness state of the working surface can be characterized in terms of image pixel features and light scattering features. The morphological contour fluctuation depth distribution attribute information and the morphological contour fluctuation interval distribution attribute information are also integrated into the plane space corresponding to the working surface, thereby generating a roughness plane distribution model of the working surface, so that the roughness plane distribution model can comprehensively and accurately characterize the roughness of the working surface.

[0090] See also Figure 2 As shown, an embodiment of the present application provides a working surface roughness detection system based on machine vision. The working surface roughness detection system based on machine vision includes:

[0091] A first visual recognition module is used to obtain a first global image of the working surface under the illumination of the reference illumination light, analyze the first global image, and obtain image shadow distribution feature information of the first global image;

[0092] An abnormal lighting area recognition module is used to determine the abnormal lighting sub-area inside the lighting space formed by the reference lighting light based on the image shadow distribution feature information;

[0093] an illumination light adjustment module, used for adjusting the illumination state of the additional illumination light on the work surface based on the attribute information of the abnormal illumination sub-region;

[0094] A second visual recognition module is used to obtain a second global image of the working surface under the combined illumination of the reference illumination light and the additional illumination light; and analyze the second global image to obtain texture feature information and scattered light feature information of the working surface;

[0095] A surface morphology analysis module, used to determine first morphology distribution attribute information of the working surface based on the texture feature information; and to determine second morphology distribution attribute information of the working surface based on the scattered light feature information;

[0096] The surface roughness distribution characterization module is used to generate a roughness plane distribution model of the working surface based on the first morphology distribution attribute information and the second morphology distribution attribute information.

[0097] The beneficial effects of the above embodiments are as follows: the machine vision-based working surface roughness detection system analyzes the first global image of the working surface under the illumination of the reference illumination light to obtain the image shadow distribution feature information, thereby determining the abnormal illumination sub-area within the illumination space formed by the reference illumination light, and accurately locating the area of ​​the working surface that is not fully and effectively illuminated; based on the abnormal illumination sub-area, the illumination state of the additional illumination light on the working surface is adjusted to achieve global uniform illumination of the working surface; the second global image of the working surface under the common illumination of the reference illumination light and the additional illumination light is analyzed to obtain the texture feature information and scattered light feature information of the working surface, and characterize the morphology of the working surface at the image vision and light scattering levels; based on the texture feature information and the scattered light feature information, the different morphology distribution attribute information of the working surface is determined respectively, thereby integrating the roughness plane distribution model of the working surface, improving the accuracy of the roughness detection of the working surface, and providing a full-range basis for the refined processing of the working surface.

[0098] In another embodiment, the first visual recognition module is used to obtain a first global image of the work surface under the reference illumination light, analyze the first global image, and obtain the image shadow distribution feature information of the first global image, including:

[0099] Scanning and photographing a working surface under the illumination of a reference illumination light to obtain a first global image of the working surface; performing image brightness recognition on the first global image to obtain image shadow distribution feature information of the first global image;

[0100] The abnormal lighting area identification module is used to determine the abnormal lighting sub-area inside the lighting space formed by the reference lighting light based on the image shadow distribution feature information, including:

[0101] Based on the image shadow distribution characteristic information, the image shadow area distribution range position of the first global image is determined; the image shadow area distribution range position is mapped to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area within the illumination space.

[0102] The beneficial effect of the above embodiment is that when the working surface of the precision parts is unevenly illuminated during the visual shooting process, the image of the working surface obtained will have obvious dark areas, so that the image of the working surface corresponding to the obvious dark areas cannot be identified and analyzed. For this reason, the working surface under the irradiation of the reference illumination light is scanned and photographed to obtain a first global image of the working surface; wherein the reference illumination light may be, but is not limited to, a collimated illumination light with a uniform cross-sectional light intensity distribution, so that the first global image can fully reflect the light intensity distribution of the working surface under the irradiation of the reference illumination light for characterization. The first global image is analyzed for the screen brightness value to obtain the screen shadow distribution feature information of the first global image; if the average brightness of a certain screen area of ​​the first global image is less than the preset brightness threshold, the corresponding screen area is determined as the screen shadow area, so that the distribution of all screen shadow areas on the first global image is fully identified and calibrated. In addition, based on the boundary feature information of the shadow distribution of the picture, the distribution range position of the shadow area of ​​the first global image is determined, and the distribution range position of the shadow area is mapped to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area inside the illumination space, so that the abnormal illumination sub-area corresponds one-to-one with the shadow area of ​​the picture, and subsequently the shadow area of ​​the picture can be directionally eliminated by performing fill light illumination in the abnormal illumination sub-area.

[0103] In another embodiment, the illumination light adjustment module is used to adjust the illumination state of the additional illumination light on the work surface based on the attribute information of the abnormal illumination sub-region, including:

[0104] Based on the distribution position information and light intensity information of the abnormal illumination sub-area, adjusting the irradiation position and irradiation intensity of the additional illumination light irradiated on the work surface;

[0105] The second visual recognition module is used to obtain a second global image of the working surface under the common illumination of the reference illumination light and the additional illumination light; and analyze the second global image to obtain texture feature information and scattered light feature information of the working surface, including:

[0106] Performing binocular photography on the working surface under the illumination of the reference illumination light and the additional illumination light to obtain a global binocular image of the working surface as the second global image; and generating a corresponding global three-dimensional image based on the binocular parallax of the global binocular image;

[0107] Pixel texture recognition is performed on the global three-dimensional image to obtain texture depth and texture direction feature information of the working surface; background light distribution extraction processing is performed on the global three-dimensional image to obtain a background light distribution image of the global three-dimensional image; light intensity distribution uniformity recognition is performed on the background light distribution image to obtain scattered light distribution feature information of the working surface.

[0108] The beneficial effect of the above embodiment is that the corresponding area of ​​the work surface is supplemented with light based on the distribution position information and light intensity information of the abnormal illumination sub-area, that is, the illumination position and light intensity of the additional illumination light are illuminated on the work surface, so that the shadow elimination illumination of the work surface can be accurately directed to ensure that the work surface is uniformly illuminated as a whole. In addition, the work surface under the common illumination of the reference illumination light and the additional illumination light is photographed and converted by binoculars to obtain a global three-dimensional image of the work surface, thereby performing a three-dimensional characterization of the morphological contour of the work surface. Then, the global three-dimensional image is subjected to pixel texture recognition, background light distribution extraction processing, and light intensity distribution uniformity recognition to obtain the texture depth and texture direction feature information of the work surface and the scattered light distribution feature information, and the surface undulation and roughness state of the work surface is characterized from the two aspects of image pixel features and light scattering features.

[0109] In another embodiment, the surface morphology analysis module is used to determine first morphology distribution attribute information of the working surface based on the texture feature information; and to determine second morphology distribution attribute information of the working surface based on the scattered light feature information, including:

[0110] Determine the depth distribution attribute information of the topographic profile undulation of the working surface based on the texture depth and texture direction characteristic information of the working surface; determine the interval distribution attribute information of the topographic profile undulation of the working surface based on the scattered light distribution characteristic information of the working surface;

[0111] The surface roughness distribution characterization module is used to generate a roughness plane distribution model of the working surface based on the first morphology distribution attribute information and the second morphology distribution attribute information, including:

[0112] The depth distribution attribute information of the topographic profile fluctuation and the interval distribution attribute information of the topographic profile fluctuation are integrated into the plane space corresponding to the working surface, so as to generate a roughness plane distribution model of the working surface.

[0113] The beneficial effect of the above-mentioned embodiment is that the morphological fluctuation of the working surface is the direct cause of the roughness of the working surface, and the morphological fluctuation of the working surface is reflected as texture depth and texture direction at the image level. At the same time, when the light is irradiated to the working surface, the corresponding rough morphology will scatter the light (the scattering characteristics of the light are related to the morphological contour fluctuation interval of the working surface), so that the background light of the image of the working surface will present corresponding light scattering characteristics. For this reason, the texture depth and texture direction feature information of the working surface are used to determine the morphological contour fluctuation depth distribution attribute information of the working surface; based on the scattered light distribution feature information of the working surface, the morphological contour fluctuation interval distribution attribute information of the working surface is determined, so that the surface undulation roughness state of the working surface can be characterized in terms of image pixel features and light scattering features. The morphological contour fluctuation depth distribution attribute information and the morphological contour fluctuation interval distribution attribute information are also integrated into the plane space corresponding to the working surface, thereby generating a roughness plane distribution model of the working surface, so that the roughness plane distribution model can comprehensively and accurately characterize the roughness of the working surface.

[0114] In general, the machine vision-based working surface roughness detection method and system analyze the first global image of the working surface under the illumination of the reference illumination light to obtain the image shadow distribution feature information, thereby determining the abnormal illumination sub-area within the illumination space formed by the reference illumination light, and accurately locating the area of ​​the working surface that is not fully and effectively illuminated; based on the abnormal illumination sub-area, adjust the illumination state of the additional illumination light on the working surface to achieve global uniform illumination of the working surface; analyze the second global image of the working surface under the common illumination of the reference illumination light and the additional illumination light to obtain the texture feature information and scattered light feature information of the working surface, and characterize the morphology of the working surface at the image vision and light scattering levels; based on the texture feature information and the scattered light feature information, respectively determine the different morphology distribution attribute information of the working surface, thereby integrating the roughness plane distribution model of the working surface, improving the accuracy of the roughness detection of the working surface, and providing a comprehensive basis for the refined processing of the working surface.

[0115] The above is only a specific implementation of the present invention, and any other improvements made based on the concept of the present invention are considered to be within the protection scope of the present invention.

Claims

1. A method for detecting the roughness of a working surface based on machine vision, characterized in that: include: Acquire a first global image of the working surface under the illumination of the reference illumination light, analyze the first global image, and obtain image shadow distribution feature information of the first global image; Based on the picture shadow distribution characteristic information, determining an abnormal illumination sub-region inside the illumination space formed by the reference illumination light; Based on the attribute information of the abnormal illumination sub-region, adjusting the illumination state of the additional illumination light on the working surface; acquiring a second global image of the working surface under the illumination of the reference illumination light and the additional illumination light; analyzing the second global image to obtain texture feature information and scattered light feature information of the working surface; Based on the texture feature information, first morphology distribution attribute information of the working surface is determined; based on the scattered light feature information, second morphology distribution attribute information of the working surface is determined; based on the first morphology distribution attribute information and the second morphology distribution attribute information, a roughness plane distribution model of the working surface is generated.

2. The method for detecting working surface roughness based on machine vision according to claim 1, characterized in that: Acquire a first global image of the working surface under the illumination of the reference illumination light, analyze the first global image, and obtain image shadow distribution feature information of the first global image; Determining an abnormal illumination sub-region within the illumination space formed by the reference illumination light based on the picture shadow distribution characteristic information includes: Scanning and photographing a working surface under the illumination of a reference illumination light to obtain a first global image of the working surface; Performing screen brightness recognition on the first global image to obtain screen shadow distribution feature information of the first global image; Based on the picture shadow distribution characteristic information, determine the picture shadow area distribution range position of the first global image; map the picture shadow area distribution range position to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area inside the illumination space.

3. The method for detecting working surface roughness based on machine vision according to claim 1, characterized in that: Based on the attribute information of the abnormal illumination sub-region, adjusting the illumination state of the additional illumination light on the working surface; acquiring a second global image of the working surface under the illumination of the reference illumination light and the additional illumination light; Analyzing the second global image to obtain texture feature information and scattered light feature information of the working surface includes: Based on the distribution position information and light intensity information of the abnormal illumination sub-area, adjusting the irradiation position and irradiation intensity of the additional illumination light irradiated on the work surface; Performing binocular photography on the working surface under the illumination of the reference illumination light and the additional illumination light to obtain a global binocular image of the working surface as the second global image; and generating a corresponding global three-dimensional image based on the binocular parallax of the global binocular image; Perform pixel texture recognition on the global three-dimensional image to obtain texture depth and texture direction feature information of the working surface; perform background light distribution extraction processing on the global three-dimensional image to obtain a background light distribution image of the global three-dimensional image; perform light intensity distribution uniformity recognition on the background light distribution image to obtain scattered light distribution feature information of the working surface.

4. The method for detecting working surface roughness based on machine vision according to claim 3, characterized in that: Based on the distribution position information and the light intensity information of the abnormal illumination sub-area, adjusting the irradiation position and the irradiation intensity of the additional illumination light irradiated on the work surface includes: Step S1, using the following formula (1), according to the distribution position information and light intensity information of the abnormal illumination sub-area, determine the irradiation position of the additional illumination light on the working surface, In the above formula (1), (X0, Y0) represents the coordinate point of the irradiation position of the working surface irradiated with the additional illumination light; Indicates coordinate points Coordinate point and coordinate points The coordinate point of the hammer point with the line of the connecting line as a perpendicular line; [X(k), Y(k)] represents the kth abnormal lighting sub-area position coordinate point in the distribution position information of the abnormal lighting sub-area; n represents the total number of abnormal lighting sub-area position coordinate points in the distribution position information of the abnormal lighting sub-area; It means to put the value of k from 1 to n into the brackets to get the maximum value in the brackets; It means to put the value of k from 1 to n into the brackets to get the minimum value in the brackets; Step S2, using the following formula (2), according to the distribution position information and light intensity information of the abnormal illumination sub-area, determine the illumination intensity of the additional illumination light irradiated on the work surface, In the above formula (2), E represents the intensity of the additional illumination light irradiated on the working surface; E q represents the light intensity of the abnormal illumination sub-area; L{[X(k), Y(k)], (X0, Y0)} represents the distance value between the coordinate point [X(k), Y(k)] and the coordinate point (X0, Y0); E0 represents the minimum light intensity of a circle with a unit distance of illumination diameter; D0 represents the unit distance; Step S3, using the following formula (3), according to the intensity of the additional illumination light irradiated on the work surface and the duration of the additional illumination light, whether it is necessary to switch from the currently used additional illumination light to other additional illumination lights, In the above formula (3), Z2 represents the judgment value for switching from the currently used additional illumination light to other additional illumination lights; T0 represents the duration of continuous illumination at the minimum light intensity of a circle with a unit distance of illumination diameter; t represents the current moment; t0 represents the moment when the additional illumination light starts irradiating; T h Indicates the switching time of the additional lighting; If Z2=1, it is necessary to switch from the currently used additional illumination light to other additional illumination lights; if Z2=0, it is not necessary to switch from the currently used additional illumination light to other additional illumination lights.

5. The method for detecting working surface roughness based on machine vision according to claim 1, characterized in that: Based on the texture feature information, determining first morphology distribution attribute information of the working surface; based on the scattered light feature information, determining second morphology distribution attribute information of the working surface; Generating a roughness plane distribution model of the working surface based on the first morphology distribution attribute information and the second morphology distribution attribute information, comprising: Determine the depth distribution attribute information of the topographic profile undulation of the working surface based on the texture depth and texture direction characteristic information of the working surface; determine the interval distribution attribute information of the topographic profile undulation of the working surface based on the scattered light distribution characteristic information of the working surface; The depth distribution attribute information of the topography profile fluctuation and the interval distribution attribute information of the topography profile fluctuation are integrated into the plane space corresponding to the working surface, so as to generate a roughness plane distribution model of the working surface.

6. A working surface roughness detection system based on machine vision, characterized in that: include: A first visual recognition module is used to obtain a first global image of the work surface under the illumination of a reference illumination light, analyze the first global image, and obtain image shadow distribution feature information of the first global image; An abnormal lighting area identification module, used to determine the abnormal lighting sub-area inside the lighting space formed by the reference lighting light based on the image shadow distribution feature information; an illumination light adjustment module, configured to adjust the illumination state of the additional illumination light on the work surface based on the attribute information of the abnormal illumination sub-region; A second visual recognition module is used to obtain a second global image of the working surface under the combined illumination of the reference illumination light and the additional illumination light; and analyze the second global image to obtain texture feature information and scattered light feature information of the working surface; A surface morphology analysis module, configured to determine first morphology distribution attribute information of the working surface based on the texture feature information; and to determine second morphology distribution attribute information of the working surface based on the scattered light feature information; A surface roughness distribution characterization module is used to generate a roughness plane distribution model of the working surface based on the first morphology distribution attribute information and the second morphology distribution attribute information.

7. The machine vision-based working surface roughness detection system according to claim 6, characterized in that: The first visual recognition module is used to obtain a first global image of the work surface under the illumination of the reference illumination light, analyze the first global image, and obtain the image shadow distribution feature information of the first global image, including: Scanning and photographing a work surface under the illumination of a reference illumination light to obtain a first global image of the work surface; performing image brightness recognition on the first global image to obtain image shadow distribution feature information of the first global image; The abnormal lighting area identification module is used to determine the abnormal lighting sub-area inside the lighting space formed by the reference lighting light based on the picture shadow distribution feature information, including: Based on the picture shadow distribution characteristic information, determine the picture shadow area distribution range position of the first global image; map the picture shadow area distribution range position to the illumination space formed by the reference illumination light, so as to determine the abnormal illumination sub-area inside the illumination space.

8. The machine vision-based working surface roughness detection system according to claim 6, characterized in that: The illumination light adjustment module is used to adjust the illumination state of the additional illumination light on the work surface based on the attribute information of the abnormal illumination sub-area, including: Based on the distribution position information and light intensity information of the abnormal illumination sub-area, adjusting the irradiation position and irradiation intensity of the additional illumination light irradiated on the work surface; The second visual recognition module is used to obtain a second global image of the working surface under the combined illumination of the reference illumination light and the additional illumination light; and analyze the second global image to obtain texture feature information and scattered light feature information of the working surface, including: Performing binocular photography on the working surface under the illumination of the reference illumination light and the additional illumination light to obtain a global binocular image of the working surface as the second global image; and generating a corresponding global three-dimensional image based on the binocular parallax of the global binocular image; Perform pixel texture recognition on the global three-dimensional image to obtain texture depth and texture direction feature information of the working surface; perform background light distribution extraction processing on the global three-dimensional image to obtain a background light distribution image of the global three-dimensional image; perform light intensity distribution uniformity recognition on the background light distribution image to obtain scattered light distribution feature information of the working surface.

9. The machine vision-based working surface roughness detection system according to claim 6, characterized in that: The surface morphology analysis module is used to determine first morphology distribution attribute information of the working surface based on the texture feature information; Determining second topographic distribution attribute information of the working surface based on the scattered light characteristic information includes: Determine the depth distribution attribute information of the topographic profile undulation of the working surface based on the texture depth and texture direction characteristic information of the working surface; determine the interval distribution attribute information of the topographic profile undulation of the working surface based on the scattered light distribution characteristic information of the working surface; The surface roughness distribution characterization module is used to generate a roughness plane distribution model of the working surface based on the first morphology distribution attribute information and the second morphology distribution attribute information, including: The depth distribution attribute information of the topography profile fluctuation and the interval distribution attribute information of the topography profile fluctuation are integrated into the plane space corresponding to the working surface, so as to generate a roughness plane distribution model of the working surface.

Images