Object 3D Profile Image Capture System

Abstract

We provide an object 3D profile image capture system. [Solution] The object 3D profile image capture system according to the present invention is used to find the 3D profile contour of an object to be measured by evaluating the change in focal quality of multiple slice images, and the system comprises a camera device, a moving device, and a control device. The camera device includes a lens module and an image sensor imaging module, and the relative positional relationship between the lens module and the image sensor imaging module corresponds to the focal plane on the object to be measured. The moving device is used to move the camera device along a movement path relative to the object to be measured, and the movement path is not perpendicular to the focal plane on the object to be measured. The control device is electrically connected to the camera device and the moving device, controls the movement of the moving device, and controls the camera device to acquire multiple slice images of the object to be measured as it moves along the movement path relative to the object to be measured.

Description

Technical Field

[0001] The present invention relates to an object profile image capture system and method, and more particularly to an object profile image capture system and method for analyzing a three-dimensional contour of an object profile from slice images obtained by a plurality of measurement object side focal planes inclined with respect to a movement path.

Background Art

[0002] A schematic diagram of an imaging system according to a prior art and an imaging method (see FIGS. 1 to 3) for analyzing a three-dimensional contour of an object surface from depth from focus (DFF) are shown. A method for analyzing a three-dimensional contour of an object surface from depth from focus (DFF) according to the prior art is to capture an image from a scene by a camera device, and from a series of a plurality of slice images obtained while a focal plane continuously moves on a measurement object (concept of computed tomography), by evaluating a change in the focus quality of the measurement object in each image, three-dimensional depth information of each location of the profile of the measurement object is analyzed. In a typical framework of the prior art, the movement path 80a of the image slice acquired by the camera device 10a is perpendicular to the measurement object side focal planes 0, 1, 2, 3, 4, and the focus evaluation value of the measurement object in the image has no horizontal image displacement between slices, so the analysis process is simple and clear (see FIG. 1).

[0003] As shown in FIG. 1, in the imaging method of the prior art, when the size of the measurement object exceeds the measurable range (FOV, field of view) of a single field of view of the camera, in a situation where the same resolution is maintained, it is necessary to define a plurality of measurement field regions, and the camera device 100a sequentially captures a plurality of slice images at the positions of each field region along the movement path 80a. While the camera device 100a moves to the next field region, when the camera device 100a moves along the movement path 81a, the camera device 100a is in an idle state where it does not capture an image, and the average number of frames per second of the total imaging of the entire stroke of the camera device 100a is far from the maximum imaging ability of the camera device 100a. This is like driving on a narrow alley, where you have to come to a complete stop at each entrance and then start driving again, resulting in an average speed much lower than on a highway. For example, a camera device 10a capable of capturing 180 images per second needs to capture an image of the object being inspected within each field of view, requiring 90 slice images for each field of view. For objects larger than one field of view, the entire detection area can be divided into 100 fields of view (e.g., 20 x 5 fields of view) for capturing. When switching the shooting position from one field of view to the next, the camera device must first accelerate, then decelerate until it is completely stable and stopped, and then capture the image of the next field of view. Even at the fastest movement speed of 0.5 to 1 second, capturing all 9000 images requires at least 50 seconds of full-speed shooting time and 50 to 100 seconds of movement time, and the total time for the inspection cycle reached 100 to 150 seconds. In reality, the camera was only able to utilize 1 / 2 to 1 / 3 of the camera system's maximum imaging capacity (18,000 frames / 100 seconds to 27,000 frames / 150 seconds), resulting in a low utilization rate and significant room for improvement.

[0004] This paper details a conventional method for measuring the contour of an object surface using depth of focus (DFF). For ease of understanding, Figures 2 and 3 show only five slice images 0a, 1a, 2a, 3a, and 4a of the top vertex 91f of the object 90f under test. Because the same camera device 10a has a fixed focal length, as shown in Figures 2 and 3, only slice image 2a of the top vertex 91f is a clear image, while slice images 0a, 1a, 3a, and 4a of the top vertex 91f are all blurry images, to varying degrees. Depth of focus (DFF) utilizes the fact that the features of the object image in the image change from blurry to clear and then blurry again, and analyzes the image slice index number (i.e., the relative height of the top vertex 91f) of the parts of the profile contour of the object under test that are in focus. Of the five slice images in this ratio, slice image 91f with index number 2a is the clearest.

[0005] The Laplacian filter, most commonly used to evaluate the sharpness of objects in a scene image, evaluates the change in grayscale values ​​of the feature points in the object image relative to those of adjacent pixels. This is a gradient function convolution operation, which calculates the change in the grayscale gradient at each pixel position in the image. In other words, it performs a convolution operation between the grayscale of each pixel and the grayscale of the surrounding adjacent pixels.

number

[0006] The kernel ω(i,j) is shown in Table 1 below. [Table 1]

[0007] Table 1 shows an example of convolution operation parameters for a 3x3 gradient function (convolution range values ​​a and b are both 1). If there is concern about noise in the sliced ​​image acquired by the camera device 100a, a smoother (e.g., a Gaussian filter) is first used, followed by a Laplace operation. Combining the smoother and Laplacian filter as a single filter (LoG) saves computational power. Details of the Gaussian filter are omitted here.

[0008] The convolution values ​​of the same pixels located in each slice are aggregated to form a one-dimensional array of image sharpness evaluation intensity values. From this one-dimensional array, the slice index number with the highest evaluation value is found and recorded. For more accurate depth analysis, high-precision floating-point index numbers are obtained by trend interpolation of the focus evaluation values ​​of preceding and succeeding slice images. Conventional techniques, in practice, apply DFF detection to capture images at equivalent long focal lengths belonging to the combined field of view. When capturing images on a fixed focal length plane relative to a moving camera or object under test, there is a slight variation in magnification between images where the object under test is out of focus at the off-focal plane and images where it is in focus at the focal plane, but this does not affect the index analysis of the one-dimensional array focal layer of each pixel region (e.g., 3x3, 5x5). Using a telecentric lens (virtual super-telephoto) design would provide even greater accuracy, but the cost of the lens was very high.

[0009] The matrix of slice index records for all pixels is assembled to obtain the relative depth value of the 3D profile of the object surface in the image coordinate system. Furthermore, the interval between adjacent slice planes of the object-side focal plane (the interval between adjacent layers of focal planes 0, 1, 2, 3, and 4) as the camera device 100a moves and captures images is a constant of the transformation unit from the image pixel coordinate system to the real-world coordinate system, from which the final 3D depth map of the object in the real world can be obtained.

[0010] The depth of focus (DFF) measurement system may capture images using a camera device 100a that moves along the Z-axis of a mounting platform (not shown) on which the camera device 100a is mounted, or it may capture images using a platform for the object being measured that moves along the Z-axis. Here, the Z-axis direction refers to the vertical direction. It should be noted that both relative movement imaging methods are capable of acquiring a series of slice images in which the focal point is located at different positions (relative depths) in front of and behind the object.

[0011] In recent years, some optical lenses have been modified to incorporate a variable curved liquid lens or to change to a DLP (DMD) reflective lens optical path design, and by performing an electronic simulation focusing ring function, the process of moving the camera device 100a or the object under measurement 90f has been eliminated. However, images acquired with liquid lenses or DLP (DMD) reflective lenses have the advantage of less vibration, but the nonlinear fluctuation relationship between the object distance (p) and image distance (q) of the lens, which is essentially a double curve (1 / p + 1 / q = 1 / f), results in a large variation in the accuracy of the analysis depending on the control of the imaging distance. Therefore, magnification correction is required to change the magnification ratio of the slice image acquisition in steps (compared to the disclosure in paragraph

[0008] ), and the subsequent image processing process has become more complex. In addition, it is necessary to capture multiple slice images along the vertical direction (Z axis) at a fixed point (horizontally stationary), but it was not possible to completely escape the fate of low utilization rate disclosed in paragraph

[0003] . [Overview of the project] [Problems that the invention aims to solve]

[0012] This invention has been made in view of the above-mentioned conventional problems. To solve the above problems, one of the objectives of this invention is to provide an object profile image capture system that analyzes the 3D contour profile of an object to be measured using slice images acquired from multiple focal planes on the object side that are inclined (non-parallel) with respect to the movement path of a camera device.

[0013] One of the objectives of this invention is to provide an object profile image capture method that analyzes the 3D contour profile of an object by using slice images acquired from multiple focal planes on the object side that are inclined (non-parallel) to the movement path of a camera device. [Means for solving the problem]

[0014] To achieve the above objective, an object profile image capture system and method in one aspect of the present invention analyzes the three-dimensional profile contour of an object by evaluating the change in the focus quality of slice images acquired at multiple object-side focal planes inclined with respect to the movement path of a camera device. The system comprises a camera device, a moving device, and a control device. The camera device includes a lens module and an image sensor imaging module, and the relative positional relationship between the lens module and the image sensor imaging module corresponds to the object-side focal plane. The moving device is used to move the camera device along a movement path relative to the object, and the movement path is not perpendicular to the object-side focal plane. The control device is electrically connected to the camera device and the moving device, controls the movement of the moving device, and controls the camera device to acquire multiple slice images of the object as it moves along the movement path relative to the object.

[0015] The present invention utilizes the characteristic of the object-side focal plane that is not perpendicular to the movement path of the camera device, and the camera device according to the present invention can bring the imaging speed of the camera device according to the present invention close to the maximum detection imaging speed by only capturing "continuous" images while moving laterally (for example, along the X-axis of the mounting mobile platform on which the camera device is mounted, or operating synchronously to interpolate along the path on the XZ plane of the mounting mobile platform on which the camera device is mounted). Taking the proposal disclosed in paragraph

[0003] of the prior art as an example, the present invention captures images in about 51 to 52 seconds (including buffers for acceleration and deceleration before and after), the idle time when moving in a different column (for example, along the Y-axis) is 2.5 to 5 seconds, the total time is less than 57 seconds, and the inspection cycle time is less than 40 to 50 percent of that of the prior art method, and the present invention is equivalent to increasing the detection capability on the production line by 2 to 2.5 times or more, and increases the imaging utilization rate of the camera device. The prior art camera device has the drawback that it idles without capturing images while moving in multiple field of view areas, which reduces the imaging utilization rate and detection capability.

[0016] The following information will become clear from the description in the specification and drawings described later.

Brief Description of the Drawings

[0017] [Figure 1] It is a schematic diagram showing a conventional image capture system. [Figure 2] It is a schematic diagram showing a conventional image capture system acquiring a slice image. [Figure 3] It is a schematic diagram showing index number selection of a slice image by conventional focus analysis. [Figure 4A] It is a top view showing an object profile image capture system according to an embodiment of the present invention. [Figure 4B] It is a block diagram showing an image processing module of an object profile image capture system according to an embodiment of the present invention. [Figure 5A] It is a side schematic view showing an object profile image capture system according to the present invention capturing an image along the movement path of the first embodiment using the lens module according to the third embodiment. [Figure 5B] It is a schematic diagram showing pixel offset stack processing of a slice image according to the present invention. [Figure 6] It is a schematic diagram showing a first embodiment of a lens module applied to an object profile image capture system according to the present invention. [Figure 7] It is an explanation of the application of the Scheimpflug principle in the first embodiment of the lens module according to the present invention. [Figure 8] It is a schematic diagram showing a second embodiment of a lens module applied to an object profile image capture system according to the present invention. [Figure 9] It is a schematic diagram showing a third embodiment of a lens module applied to an object profile image capture system according to the present invention. [Figure 10] It is a side schematic view showing an object profile image capture system according to the present invention capturing an image along the movement path of the second embodiment using the lens module according to the second embodiment. [Figure 11] This is a flowchart showing a first embodiment of the object profile image capture method according to the present invention. [Figure 12] This is a schematic diagram showing how to acquire a slice image and perform pixel offset stacking using a lens module according to a second embodiment of the present invention. [Figure 13A] This is a schematic diagram showing the image to be evaluated after the pixel offset stacking process of the sliced ​​image has been completed. [Figure 13B] This is a schematic diagram showing the image to be evaluated after the pixel offset stacking process of the sliced ​​image has been completed. [Figure 14A] This schematic diagram shows a 3D depth map of a slice image obtained by a lens module according to a second embodiment of the present invention, in an object image coordinate system with reference coordinates set to the normal of the focal plane on the side of the object being measured, and in a world coordinate system with reference coordinates set to the normal of the first plane. [Figure 14B] This schematic diagram shows a 3D depth map of a slice image obtained by a lens module according to a second embodiment of the present invention, in an object image coordinate system with reference coordinates set to the normal of the focal plane on the side of the object being measured, and in a world coordinate system with reference coordinates set to the normal of the first plane. [Figure 15] This is a flowchart showing a second embodiment of the object profile image capture method according to the present invention. [Figure 16] This is a schematic diagram showing the focal plane corresponding to the keystone correction of the image by the image sensor. [Figure 17] This is a flowchart showing a third embodiment of the object profile image capture method according to the present invention. [Figure 18] This schematic diagram shows how an object point on the focal plane obtained by the lens module according to the third embodiment of the present invention corresponds to the imaging focus of the image sensor, and how it changes from a blurred state to a sharp state and then back to a blurred state. [Figure 19]This is a schematic diagram showing how to acquire a slice image and perform pixel offset stacking using a lens module according to a third embodiment of the present invention. [Figure 20] This is a schematic diagram showing that a slice image obtained by a lens module according to a third embodiment of the present invention is processed by an object profile image capture method according to the present invention and finally converted into a world coordinate system image. [Figure 21] This is a schematic diagram showing that a slice image obtained by a lens module according to a third embodiment of the present invention is processed by an object profile image capture method according to the present invention and finally converted into a world coordinate system image. [Figure 22] This is a schematic diagram showing that a slice image obtained by a lens module according to a third embodiment of the present invention is processed by an object profile image capture method according to the present invention and finally converted into a world coordinate system image. [Modes for carrying out the invention]

[0018] The present invention will be described below through embodiments of the invention, but these embodiments are not intended to limit the invention as defined in the claims. Furthermore, not all combinations of features described in the embodiments are necessarily essential to the solution of the invention.

[0019] Figure 4A is a top view showing an object profile image capture system according to one embodiment of the present invention. Figure 5A is a schematic side view showing the object profile image capture system according to the present invention capturing an image along the movement path of the first embodiment using a lens module according to the third embodiment. Figure 5B is a schematic diagram showing the pixel offset stacking process of sliced ​​images according to the present invention.

[0020] As shown in Figures 4A and 5A, in the first embodiment of the present invention, the object profile image capture system 1 according to the present invention finds the profile contour 91 of the object to be measured 90 by evaluating the change in focus quality of slice images acquired from a plurality of object-side focal planes inclined with respect to the movement direction of the camera device. The object to be measured 90 may be a circuit board, motherboard, surface mount element, such as an IC chip, resistor, capacitor, inductor, bare die, or wafer. In this embodiment, the object profile image capture system 1 according to the present invention comprises a camera device 10, a moving device 20, an object-side platform 30, and a control device 50. The camera device 10 includes a lens module 11 and an image sensor imaging module 13, and the relative positional relationship between the lens module 11 and the image sensor imaging module 13 corresponds to the object-side focal plane 15, and basically the object-side focal plane 15 is considered to be a slice plane in which the image captured by the camera device 10 is clear. In a specific embodiment of the present invention, the relative positional relationship between the lens module 11 and the image sensor imaging module 13 corresponds to the focal plane 15 on the side of the object being measured, based on the Scheimpflug Intersection Principle. However, the present invention is not limited to this embodiment.

[0021] The moving device 20 moves the camera device 10 along the movement path 80 relative to the object to be measured 90, and the movement path 80 of the moving device 20 is not perpendicular to the focal plane 15 on the object to be measured side. The control device 50 is electrically connected to the camera device 10 and the moving device 20, and the control device 50 controls the movement of the moving device 20 and controls the camera device 10 to acquire slice images of the object to be measured 90 at the focal planes 15c, 15d, and 15e of the object to be measured, respectively, as the camera device 10 moves along the movement path 80 relative to the object to be measured 90, thereby forming a plurality of slice images according to the present invention. It should be noted here that, as shown in Figure 5A, the camera device 10 in this embodiment moves along the movement path 80 relative to the object to be measured 90 in the first plane 70, and the first plane 70 in this embodiment is parallel to the object to be measured platform reference plane 35 formed by the object to be measured platform 30, that is, the XY plane in Figure 5A. Incidentally, as shown in Figures 5A and 5B, the object-side focal plane 15 has a focal plane normal 151, and the straight-line distance between adjacent object-side focal planes 15c, 15d, and 15e (in the direction along the focal plane normal 151) is defined as the slice interval thickness 880, and the distance traveled by the camera device 10 in the first plane 70, moving from the object-side focal plane 15c along the direction of the path 80 to the next object-side focal plane (for example, moving from the object-side focal plane 15d to the object-side focal plane 15e), is defined as the image interval distance 890.

[0022] According to a specific embodiment of the present invention, the imaging frequency of the camera device 10 is 1 to 1000 frames / second, and the lens module 11 may be a lens composed of one or more optical lenses, such as a microscope lens, a numerical aperture lens with a shallow depth of field, or a telecentric lens with an equivalent focal length (EFL). The image sensor imaging module 13 may be a charge-coupled device (CCD), a complementary metal-oxide-semiconductor (CMOS) device, or an indium gallium arsenide (InGaAs) device. The focal plane 15 on the side of the object to be measured is determined by the angle between the lens module 11 and the image sensor imaging module 13 used in the camera device 10, based on the Schheimfluke principle, and this is the sensor deflection angle 17(φ) shown in Figure 6.

[0023] As shown in Figures 4A and 5A, the object to be measured 90 is placed on the object to be measured platform 30, and the object to be measured platform 30 forms the object to be measured platform reference surface 35. The object to be measured platform 30 is divided into multiple camera imaging field of view (FOV) 31, making it easier for the camera device 10 to acquire slice images of the object to be measured 90. In this embodiment, the moving device 20 includes a first axis moving guide rail 21, a second axis moving guide rail 22, a third axis moving guide rail 23, and drive devices (not shown) installed on each transfer axis. This allows the camera device 10 to be adjusted to the appropriate height measurement range to meet the needs of different applications. In this embodiment, the first axis moving guide rail 21 is the X-axis moving guide rail, the second axis moving guide rail 22 is the Y-axis moving guide rail, and the third axis moving guide rail 23 is the Z-axis moving guide rail. The third-axis moving guide rail 23 moves in accordance with the first-axis moving guide rail 21 and the second-axis moving guide rail 22, and the camera device 10 moves up, down, left, right, forward, and backward above the platform 30 by a combination of synchronous movements (XZ, XY, YZ, XYZ) in which it performs oblique linear motion in various spaces and planes. In this embodiment, by installing the camera device 10 on the third-axis moving guide rail 23, the first-axis moving guide rail 21, the second-axis moving guide rail 22, and the third-axis moving guide rail 23 allow the camera device 10 to form the entire imaging range of the object platform 35 shown in Figure 4. In this embodiment, the control device 50 is a control device, processing device, or control software installed on an electronic device (e.g., a personal computer or a programmable logic controller PLC).

[0024] Here, the first plane 70 has a first plane normal 71. When the first plane 70 is the XY plane and the first plane normal 71 is the Z axis, and the camera device 10 captures images along the movement path 80 in the XY plane, it should be noted that the third axis movement guide rail 23 (Z axis) does not need to move perpendicular to the first axis movement guide rail 21 (X axis). In other words, the third axis movement guide rail 23 on which the camera device 10 is mounted only needs to move horizontally (left and right as shown in 4A and Figure 5A) in the lateral direction relative to the first axis movement guide rail 21, and imaging will be completed sequentially. As shown in Figures 4A, 5A, and 5B, during the process of the camera device 10 moving along the movement path 80, the camera device 10 is actually in different positions (displacement interval distance 890). Therefore, as the camera device 10 moves continuously and takes images, the camera device 10 acquires slice images corresponding to each of the object-side focal planes 15c, 15d, and 15e at each imaging position. The multiple slice images acquired by the camera device 10 are images of the object 90 that change from blurry to sharp and then blurry again, and are provided for subsequent analysis of the 3D profile contour of the object 90.

[0025] Furthermore, according to a specific embodiment of the present invention, as shown in Figure 5A, the focal plane deflection angle 19(ω) is formed by the focal plane normal 151 of the focal plane 15 on the object-side, and the first plane normal 71, and the angular range of the focal plane deflection angle 19(ω) is between 0.1° and 60°, 1° and 10°, 0.5° and 20°, or between 0.1° and 45°. As shown in Figure 5A, since the focal plane 15 on the object-side according to the present invention is not parallel to the first plane 70, the slice image acquired by the camera device 10 taking continuous images along the movement path 80 on the first plane 70 is actually sliced ​​diagonally, and this diagonal slice image can show even more profile contour features of the side surface of the object 90. Furthermore, an advantage of the present invention is that the focal plane deviation angle 19(ω) can be adjusted to correspond to differences in the required depth measurement range of the object to be measured 90. For example, if the object to be measured 90 is a circuit board, motherboard, or surface mount element with a high element height, the focal plane deviation angle 19(ω) can be adjusted to be larger, for example, 25°, 45°, or 60°, to obtain a clearer and more detailed contour of the side surface of the object to be measured 90. If the object to be measured 90 is an element with a low element height, such as an IC chip, resistor, capacitor, inductor, bare die, or wafer, the focal plane deviation angle 19(ω) can be adjusted to be smaller, for example, 1°, 5°, or 10°, to obtain depth measurement information inside a hole with a high aspect ratio of the object to be measured 90.

[0026] Here, using an image sensor imaging module 13 with a SONY CMOS IMX535, 4Kx3K resolution, HxV dimensions of 11.2mm x 8.2mm, 12 million pixels, and a pixel size of 2.74um as an example, if the object profile image capture system 1 according to the present invention is an optical detection system used in surface mount (SMT / SMD) manufacturing lines, then such a system has a wide height measurement range for the object being measured. Based on trigonometry, if the focal plane declination angle 19(ω) is 30° and the optical magnification of the camera device 10 is 0.5X, the measurable height measurement range will be approximately 11.2mm (Sin30° / 0.5X=1). When the object profile image capture system 1 according to the present invention is used in an optical detection system for the wafer surface, the height measurement range for objects on the manufacturing line is narrow. Therefore, if the focal plane declination angle 19(ω) is 2.56° and the optical magnification of the camera device 10 is 5X (for example, the objective lens of a microscope), the obtainable height measurement range will be approximately 100um. In this way, the height measurement range can be adjusted with approximately 1120 times more flexibility.

[0027] Please also refer to Figures 4A and 5A below. Figures 6 and 7 show a schematic diagram of a first embodiment applied to the lens module according to the present invention and an explanation of how the Schheimfluke principle is applied to this embodiment.

[0028] As shown in Figure 6, in the first embodiment of the lens module 11, the movement path 80 refers to the path along which the camera device 10 moves in the horizontal direction (X-axis) parallel to the first plane 70 (XY plane) composed of the first axis movement guide rail 21 (X-axis) and the second axis movement guide rail 22 (Y-axis). The lens optical central axis 111 formed by the lens module 11 according to this embodiment is perpendicular to the movement path 80, and the angular range of the sensor deviation angle 17 (φ) formed by the image sensor imaging plane normal 131 formed by the image sensor imaging module 13 and the lens optical central axis 111 is between 0.1° and 60°. When the camera device 10 moves along the movement path 80 in the first plane 70, the images captured by the image sensor imaging module 13 at the focal planes 15a, 15b, 15c, 15d, and 15e on the object-to-measure side are slice images 800a, 800b, 800c, 800d, and 800e, respectively, as shown in Figure 6. The angle of the tilted focal plane is determined by a correspondence that conforms to the Schheimfluke principle, as shown in Figure 7. It should be noted that the sensor deviation angle 17(φ) of the present invention should be adjusted to correspond to differences in the shape and size of the object to be detected in practical applications. If the object to be measured 90 is a circuit board, motherboard, or surface mount element with a wide height measurement range, the sensor deviation angle 17(φ) should be adjusted to be larger (for example, 25°, 45°, 60°, etc.). If the object to be measured 90 is an element with a narrow height measurement range, such as an IC chip, resistor, capacitor, inductor, bare die, or wafer, the sensor deflection angle 17(φ) is adjusted to a small value such as 1°, 5°, or 10° to suit the measurement of a micro-level high-resolution 3D profile and the depth of the hole bottom.

[0029] Please also refer to Figures 4A and 5A below, as Figure 8 is a schematic diagram showing a second embodiment applied to the lens module according to the present invention.

[0030] As shown in Figure 8, in the second embodiment of the lens module 11, the movement path 80 refers to the path along which the camera device 10 moves in the horizontal direction (X-axis) in the first plane 70 (XY plane). The lens optical central axis 111 formed by the lens module 11a according to this embodiment is not perpendicular to the movement path 80, while the image sensor imaging module 13 is perpendicular to the lens optical central axis 111. That is, the image sensor imaging plane normal 131 of the image sensor imaging module 13 is parallel to the lens optical central axis 111. Furthermore, the angular range of the lens declination angle 18(θ) formed by the lens optical central axis 111 and the movement path 80 according to this embodiment is between 0.1° and 60°, and it should be noted that the camera device lens 11a according to this embodiment is a common conventional optical lens module. Because the lens optical central axis 111 is not perpendicular to the movement path 80, when the lens module 11a moves along the movement path 80 in the first plane 70, the images captured by the image sensor imaging module 13 at the focal planes 15a, 15b, 15c, 15d, and 15e on the object-to-measure side are the slice images 800a, 800b, 800c, 800d, and 800e shown in Figure 8, respectively. It should be noted that the lens declination angle 18(θ) formed by the lens optical central axis 111 and the movement path 80 in this embodiment is adjusted according to the differences in the shape and size of the object 90 to be measured. In applications, if the object 90 to be measured is a circuit board, motherboard, or surface mount element with a high element height, the lens declination angle 18(θ) can be adjusted to be larger (for example, 25°, 45°, 60°, etc.), and contour measurement of the side surface of the object 90 to be measured can be obtained (this is not possible with the conventional DFF method in which the lens centerline is perpendicular to the Z axis and images are captured vertically up and down). If the object to be measured 90 is an element with a narrow height measurement range, such as an IC chip, resistor, capacitor, inductor, bare die, or wafer, the lens declination angle 18(θ) is adjusted to be small, for example, to 1°, 5°, or 10°, to enable measurement of the external side of the object or the inner wall (1° Tilt) of a hole with an aspect ratio of 57.3:1.

[0031] Please also refer to Figures 4A and 5A below, as Figure 9 is a schematic diagram showing a third embodiment applied to the lens module according to the present invention.

[0032] As shown in Figure 9, in the third embodiment of the lens module 11, the movement path 80 refers to the path along which the camera device 10 moves in the horizontal direction (X-axis) in the first plane 70 (XY plane). The lens optical central axis 111 formed by the lens module 11b according to this embodiment is not perpendicular to the movement path 80, the image sensor imaging module 13 is parallel to the movement path 80, and the sensor declination angle 17 (φ) is equal to the lens declination angle 18 (θ). The angular ranges of φ and θ are both between 0.1° and 60°. When the camera device 10b moves along the movement path 80 in the first plane 70, the images captured by the image sensor imaging module 13 at the focal planes 15a, 15b, 15c, 15d, and 15e on the object side are the slice images 800a, 800b, 800c, 800d, and 800e shown in Figure 9, respectively. In this embodiment, the sensor deflection angle 17(φ) and lens deflection angle 18(θ) are adjusted according to the differences in the object 90 being measured. In practical applications, if the object 90 being measured is a circuit board, motherboard, or a surface mount element with a high element height, it should be noted that the sensor deflection angle 17(φ) and lens deflection angle 18(θ) should be adjusted to be larger, for example, by adjusting φ and θ to 25°, 45°, 60°, etc., in order to obtain a clearer and finer contour of the side of the object 90 being measured. If the object 90 being measured is an element with a low element height, such as an IC chip, resistor, capacitor, inductor, bare die, or wafer, the sensor deflection angle 17(φ) and lens deflection angle 18(θ) should be adjusted to be smaller, for example, by adjusting φ and θ to 1°, 5°, 10°, etc., in order to make the measurable aspect ratio 57:1 or even larger and deeper (similar to the lens module 11a in the second embodiment).

[0033] Please also refer to Figures 4A and 5A below. Figure 10 is a schematic side view showing how the object profile image capture system according to the present invention captures images along the movement path of the second embodiment using the lens module according to the second embodiment.

[0034] As shown in Figures 4A, 5A, and 10, the second embodiment of the movement paths 80b and 80c differs from the first embodiment of the movement path 80 in that, when the camera device 10 according to the second embodiment takes images along the movement paths 80b and 80c, the third axis movement guide rail 23 (Z axis) moves horizontally (left and right) in the lateral direction relative to the first axis movement guide rail 21 (X axis), and at the same time, the third axis movement guide rail 23 moves perpendicularly to the first axis movement guide rail 21, and the third axis movement guide rail 23 on which the camera device 10 is mounted moves diagonally (diagonally upward, diagonally downward) relative to the first axis movement guide rail 21. In other words, the movement path 80 of this embodiment has an upward path (shown in movement path 80c) and a downward path (shown in movement path 80b). In this case, the first plane 70a is the moving inclined surface (XZ) of the camera device 10, which is composed of the first axis movement guide rail 21 and the third axis movement guide rail 23. Specifically, as can be seen from the side view shown in Figure 10, the normal vector 131 of the image sensor imaging plane of the camera device 10b is parallel to the lens optical central axis 111, and the normal vector 131 of the image sensor imaging plane and the lens optical central axis 111 are perpendicular to the reference plane 35 of the object under measurement platform. Also, in the viewing angle shown in Figure 10, the camera device 10b as a whole has a path composed of continuously overlapping movement paths 80b and 80c perpendicular to the XZ plane of the first axis movement guide rail 21, which is a movement path similar to the English letter V (V, VV, VVV…), and the first axis movement guide rail 21 and the third axis movement guide rail 23 are configured to operate synchronously in a diagonal straight path (+XZ, +X+Z). The object-side focal planes 15a, 15b, 15c, 15d, 15e, 15f, 15g, 15h, and 15i, to which the movement paths 80b and 80c are combined, are parallel to the object-side platform reference plane 35. In this case, the object profile image capture system 1 according to this embodiment uses the camera device 10a shown in Figure 8, and the lens module 11a has the normal vector 131 of the image sensor imaging plane and the optical central axis 111 of the lens parallel to each other. It should be noted that this lens module 11a is a common conventional optical lens module.

[0035] As shown in Figure 10, in this embodiment, since the focal plane 15 on the object to be measured is parallel to the reference plane 35 of the object to be measured platform (the same as the DFF imaging focal plane in the prior art), as shown in Figure 10, in an embodiment where the first axis moving guide rail 21 and the third axis moving guide rail 23 operate synchronously along diagonal straight lines (+XZ, +X+Z) moving paths 80b and 80c, when the camera device 10a moves along the moving path 80b in the direction approaching the object to be measured 90, the camera device 10 sequentially acquires slice images of the focal planes 15a, 15c, 15e, 15g, and 15i on the object to be measured. When the camera device 10a moves along the moving path 80c in the direction away from the object to be measured 90, the camera device 10a sequentially acquires slice images of the focal planes 15h, 15f, 15d, and 15b on the object to be measured again. In other words, according to a specific embodiment of the present invention, when the movement paths 80b and 80c on the first plane 70a are not parallel to the first axis movement guide rail 21, the imaging slice spacing thickness 880 of the camera device 10a along the movement paths 80b and 80c are used in an intersecting manner so that the spacing images of the V-shaped imaging paths of the movement paths 80b and 80c are combined. As a result, the camera device 10a makes the overall actual slice spacing thickness for capturing images along the movement paths 80b and 80c half of the slice spacing thickness 880, and doubles the overall slice resolution for capturing images along the movement paths 80b and 80c.

[0036] For example, as shown in Figure 10, when the camera device 10a moves along movement paths 80b and 80c in the first plane 70a to capture images, if the camera device 10a is on movement paths 80b and 80c, the camera device 10a performs interval imaging (for example, first acquiring layers with index numbers 1, 3, and 5 on movement path 80b, and then acquiring layers 2, 4, and 6 on movement path 80c), and then sequentially stacks the original 8 columns again (stacking in the order of index numbers 1, 2, 3, 4, 5, and 6). The depth range along movement paths 80b and 80c in this embodiment is the same as in the prior art and is determined based on the specification of the working distance of the front end of the lens. Furthermore, in this embodiment, the paths of movement paths 80b and 80c are determined by determining the hypotenuse of a right triangle composed of half the width of the horizontal field of view of the camera device 10a and the depth of the Z-axis moving slice.

[0037] Furthermore, when the camera device 10a changes from the movement path 80b (upward path) to the movement path 80c (downward path), or from the movement path 80c (downward path) to the movement path 80b (upward path), the camera device 10a needs to be shifted by half a slice thickness 880 above (or below) the beginning of movement along the movement path 80b or movement path 80c. By overlapping and stacking a series of slice images taken along the movement path 80b and movement path 80c, subsequent focus analysis is performed. However, the present invention is not limited to the embodiments described above. As shown in Figure 10, when the camera device 10a has m unit imaging ranges 31 along the direction of movement, and the camera device 10a moves along a V-shaped path (movement paths 80b and 80c), the present invention is applied when the camera device 10a completes the V-shaped path, that is, after the camera device 10a completes the ascending path (movement path 80b) and the ascending path (movement path 80c), the camera device 10a moves from the nth unit imaging range 31 to the (n+1)th unit imaging range 31 of the m unit imaging ranges 31. n and m are natural numbers and m > n.

[0038] Please also refer to Figures 4A, 5A, 5B, and Figures 6 through 10 below. Figure 4B is a block diagram showing the image processing module of the object profile image capture system according to the present invention.

[0039] The object profile image capture system 1 according to the present invention further comprises an image processing module 60 to which signals are connected to a control device 50 to receive a plurality of slice images 800 (see Figures 4A and 4B), and the image processing module 60 includes an evaluation module 61, an image space transformation module 62, and a matrix transformation module 63. When the camera device 10 moves along a movement path 80 with respect to the object to be measured 90 at a magnification level and acquires a plurality of slice images 800 of the object to be measured 90, and there is an image spacing distance 890 between each adjacent slice image 800, the image processing module 60 performs pixel offset stacking on the plurality of slice images 800 based on a combination of parameters such as the magnification level, the image spacing distance 890, and the focal plane declination angle 19, in order to generate a plurality of evaluation target images. The evaluation module 61 evaluates the focus quality of multiple images to be evaluated using a Laplacian filter focus evaluation calculation process for depth of field (DFF), and completes a 3D depth map 710 in an image coordinate system with the normal of the focal plane 151 on the object being measured as the reference coordinate. The image space transformation module 62 transforms the 3D depth map 710 of the image coordinate system of an object in an image coordinate system with the normal of the focal plane 151 on the object being measured as the reference coordinate into a 3D depth map 300 in a world coordinate system with the normal of the first plane 71 as the reference, based on the focal plane deviation and the image spacing distance.

[0040] It should be noted that, in the case of the lens module 11 according to the first embodiment, the image sensor imaging module 13 is not parallel to the focal plane 15 on the object side. Therefore, in the case of the lens module 11 according to the first embodiment of the present invention, the matrix transformation module 63 performs a geometric deformation matrix transformation on each of the multiple slice images 800 before the image processing module 60 performs the pixel offset stacking process on each of the multiple slice images 800. In the case of the lens module 11b according to the third embodiment of the present invention, the matrix transformation module 63 performs a geometric deformation matrix transformation on each of the multiple slice images 800 after the image coordinate system 3D depth map is completed. In the preferred embodiments of each module described above, each module is a software program, and the processing unit (not shown) or control unit 50 of the object profile image capture system 1 executes each module to achieve the function of the image processing module 60.

[0041] Please also refer to Figures 4A, 5A, and 8 below. Figure 11 is a flowchart showing a first embodiment of the object profile image capture method according to the present invention.

[0042] As shown in Figures 4A, 5A, and 11, the object profile image capture method according to the present invention is used in the object profile image capture system 1 according to the present invention, and each step of the object profile image capture method according to the present invention will be described below. As shown in Figure 11, the object profile image capture method according to the present invention includes steps S1 to S5 (see Figure 11).

[0043] Step S1: The camera device 10 moves along the movement path 80 relative to the object to be measured 90, and the movement path 80 is not perpendicular to the focal plane 15 on the object to be measured side.

[0044] As shown in Figures 4A, 5A, and 10, there are two types of methods by which the camera device 10 moves along the movement path 80 relative to the object to be measured 90. Method 1: Two embodiments in which the camera device 10 moves and captures images, and the camera device 10 moves along a first plane 70. Step S2 is performed simultaneously as the camera device 10 moves continuously along the XY plane formed by the first axis moving guide rail 21 and the second axis moving guide rail 22 (first plane 70 in Figure 5), or on a moving inclined surface composed of two mounting moving platforms, the first axis moving guide rail 21 and the third axis moving guide rail 23 (first plane 70a in Figure 10). When the camera device 10 moves along the moving path 80 relative to the object to be measured 90, the camera device 10 acquires multiple slice images 800 of the object to be measured 90. Method 2: In the two embodiments in which the camera device does not move and the object to be measured 90 is aligned with the first plane 70, step S2 is performed simultaneously when the camera device 10 moves relative to the object to be measured 90 and the object to be measured 90 moves. As the camera device 10 moves along the movement path 80 relative to the object to be measured 90, the camera device 10 acquires multiple slice images 800 of the object to be measured 90. The imaging effect of the two types of moving imaging methods described above is the same, and it should be noted that Method 2 (the object to be measured 90 moves and the camera device 10 is stationary) is used when measuring on a platform with extremely high moving and mounting accuracy where high resolution is required for images of objects on a laboratory tabletop.

[0045] The following diagrams, along with Figures 12, 13A, 13B, 14A, and 14B, illustrate schematic diagrams showing the evaluation target images generated when the slice images in steps S3 to S5 of the present invention complete the pixel offset stacking process, and illustrate the processing method for converting slice images from a 3D depth map 710 in an image coordinate system based on the normal 151 of the focal plane on the object being measured to a 3D depth map 300 in a world coordinate system based on the normal 71 of the first plane.

[0046] Step S3: Pixel offset stacking is performed on multiple slice images based on the magnification, image spacing distance, and focal plane declination to generate multiple evaluation images.

[0047] The camera device according to the present invention performs continuous imaging while moving, so the positions of the object points 92c, 92d, and 92e of the object under measurement 90 in different slice images 800, 800a, 800b, 800c, and 800d change in steps. Therefore, the amount of pixel translation is regression corrected, and the object under measurement 90 is superimposed on the positions of the slice images 800, 800a, 800b, 800c, 800d, and 800e to generate multiple evaluation target images 700 shown in Figure 13A, evaluate the subsequent depth position (changing from blurry to sharp and then blurry again), and analyze the image coordinate system depth map of the slice index number (relative height) of the sharp focal length. Specifically, the present invention corrects the pixel offset stacking process of adjacent layers in the oblique slice images 800a, 800b, 800c, 800d, and 800e based on the adjacent image spacing distance 890 and the image magnification ratio of the camera device.

[0048] As shown in Figures 5A and 12, the camera device 10b is a conventional camera. As shown in Figure 8, along the movement path 80, for example, the camera device 10b moves an equidistant distance to the left side of the drawing, and the focal plane normal 151 of the camera device 10b is parallel to the lens optical central axis 111, and the focal plane normal 151 is parallel to the image sensor imaging plane normal 131. Therefore, in this case, the lens declination angle 18 (θ, the angle between the focal plane normal 151 and the image sensor imaging plane normal 131) is equal to the focal plane declination angle 19 (ω, the angle between the focal plane normal 151 and the first plane normal 71), and the sensor declination angle 17 (φ, the angle between the lens optical central axis 111 and the image sensor imaging plane normal 131) becomes 0°. If θ=ω=10°, the magnification is 0.5X.

[0049] As shown in Figure 12, the camera device 10a captures slice images 800a, 800b, 800c, and 800d of a single object 90 at the focal planes 15a, 15b, 15c, 15d, and 15e of the object to be measured, respectively. The object points 92c, 92d, and 92e of the object 90 are displayed as 820a, 830a, and 840a, respectively, in the contour feature point image of slice image 800a, and the object points 92c, 92d, and 92e are shown in Figure 12 in slice images 800b, 800c, and 800d using the same display method as slice image 800a. For convenient display, the specific positions of the camera device 10a during movement are all indicated by the center points 16a, 16b, 16c, 16d, and 16e of the lens module 11, and the center points 16a, 16b, 16c, 16d, and 16e correspond to the image center reference points 817a, 817b, 817c, 817d, and 817e in the slice images 800a, 800b, 800c, and 800d, respectively. As shown in Figure 12, in the slice image 800a, the contour feature point image 820b and the center point 16b of object point 92c have a pixel offset T between them and the image center reference point 817b of slice image 800a. There is a pixel offset T' between the contour feature point image 820a of object point 92c in slice image 800a and the image center reference point 819a of center point 17a in slice image 800a, where T = 2T'. The reason why the pixel offsets T and T' of slice images 800a, 800b, 800c, and 800d are different is that, in this invention, slice image 800c is the center, and the image center reference point 817c of slice image 800c is the central reference point. Therefore, the further away from slice image 800c, the larger the pixel offset T becomes. If we assume that the total number of slice images is p+1, the p / 2 layer is the central slice, and the image spacing distance 890 between two adjacent slice images is q, then the pixel offset T of the p / 2+1 layer and the p / 2-1 layer will all be q, and the pixel offset T of the 1st layer and the p+1 layer will all be q × p / 2. Therefore, after the pixel offset stacking process of the slice images 800a, 800b, 800c, and 800d in Figure 12 is completed, multiple evaluation images 700 shown in Figure 13A are generated, and as shown in Figure 13A, the object point 92c has a clear image in the aligned slice image 800c'.

[0050] Step S4: The multiple images to be evaluated are evaluated using the focus evaluation calculation process of the Laplacian filter for depth of focus (DFF), and a 3D depth map of the image coordinate system is completed.

[0051] Specifically, step S4 applies the focus evaluation calculation process of the conventional depth of focus (DFF) Laplacian filter to find the image slice index number of the best focus (convolution calculation value) from each pixel of the slice image of the object under test 90 from a series of images to be evaluated 700 (finding the focal position where 820c corresponds to the object point 92c in the 800c), and completes the object image coordinate system 3D depth map 710 of the image coordinate system 200 with the object-side focal plane normal 151 as the reference coordinate (see Figure 13B). If, as shown in Figure 13A, the object point 92c has a sharp contour feature point image 820c in the aligned slice image 800c', the Laplacian filter records the index number (relative height) of the aligned slice image 800c' corresponding to the sharp contour feature point image 820c. It should be noted here that the 3D depth map 710 displays the profile contour 91 of the object under test 90. Since the technical details of the focus evaluation calculation process for the Laplacian filter of depth of focus (DFF) are conventional, we will not repeat the detailed explanation of that calculation process.

[0052] Step S5: Based on the focal plane declination and image spacing distance, convert the 3D depth map in the image coordinate system to a 3D depth map in the world coordinate system.

[0053] In detail, the 3D depth map generated in step S4 is a 3D depth map 710 generated in an image coordinate system with the normal vector 151 of the focal plane on the object being measured as the reference coordinate; therefore, further image spatial transformation is necessary (see step S5). The spatial rotation matrix is ​​Ry(ωy),

number

[0054] Please also refer to Figures 4A, 5A, 6, 7, and 9 below. Figures 15 and 16 are flowcharts showing a second embodiment of the object profile image capture method according to the present invention, and a projection coordinate transformation image processing method in which the focal plane corresponds to keystone correction by an image sensor.

[0055] In the lens modules 11 and 11b (first and third embodiments) according to the present invention, the object-side focal plane 15 and the image sensor imaging plane 13 are not parallel and have an angle between them. As a result, projection distortion occurs in the imaging of the object's focal plane in the multiple slice images 800a, 800b, 800c, 800d, and 800e acquired by the camera device 10. Therefore, it is necessary to perform spatial projection geometric transformation image processing on the slice images 800a, 800b, 800c, 800d, and 800e to accurately display the object 90 at the image position in the world coordinate system of the world coordinate system 300, which is based on the first plane normal 71.

[0056] The basic principle of thin-lens optical imaging is that the point of the object to be measured is formed when the light rays parallel to the principal point of the lens pass through the lens and then converge after passing through the principal point of the lens. By placing an image sensor at this point, a sharp image point can be obtained, and other objects in front of and behind this object to be measured gradually become blurred (see Figure 7). By adjusting the tilt angle of the image sensor at two specific non-overlapping points (and three points, plane) in the field of view, these two points (three points, plane) become sharp simultaneously. This is the Scheimphlug Intersection Principle of optical path imaging. The Scheimphlug Principle refers to drawing oblique tangents from the image plane, object plane, and lens plane when a planar object is not parallel to the image plane, and the intersection point is called the Scheimphlug intersection. By applying this principle to the technology of the present invention, the lock relationship between the lens deviation angle 17 (φ), sensor deviation angle 18 (θ), and focal plane deviation angle 19 (ω) is determined.

[0057] A feature of the present invention is that, as the camera device 10 moves, slice images (oblique slice images 800) of the inclined focal plane of the object to be measured 90 are acquired. Therefore, as the object to be measured 90 moves across the image sensor imaging plane of the image sensor imaging module 13, images are captured at different positions. Accordingly, for the lens modules 11 and 11b according to the present invention (first and third embodiments), as shown in Figure 15, it is necessary to perform a geometric matrix image coordinate transformation step (step S21: perform geometric deformation matrix transformation for each slice image 800) in the space of the image sensor corresponding to the focal plane before capturing images along the movement path and performing pixel offset stacking (pixel alignment along moving direction, step S3).

[0058] As shown in Figure 16, based on the aforementioned Scheimpflug Intersection Principle, the lens modules 11 and 11b (first and third embodiments) according to the present invention are arranged at an angle, and their respective focal plane declination angles 19(ω)≠0. As shown in Figure 16, the image sensor imaging module 13 of the lens module 11 according to the first embodiment and the lens module 11b according to the third embodiment, and the focal plane 15 of the object to be measured, are related such that one rectangular image sensor plane corresponds to a trapezoidal focal plane. An object in the world coordinate system with the first plane normal 71 as the reference coordinate is projected onto the image coordinate system with the object-side focal plane normal 151 as the reference coordinate, causing deformation. Therefore, a geometric deformation matrix transformation is performed on multiple slice images 800 to transform all trapezoidal images contained in each slice image acquired by the lens modules 11 and 11b back into rectangles in the world coordinate system. The positional pixel points a(-20.5, 14.27), b(11.055, 7.70), c(-20.5, -14.27), d(11.055, -7.70) of slice image 800, and the positional coordinates of the feature points of the object under test at focal plane 15 A(-30, 20), B(30, 20), C(-30, -20), D(30, -20) are completed by the transformation matrix shown in Figure 16, and the parameters of this matrix are obtained by the Schheimflug principle shown in Figure 7.

[0059] The following diagrams, with reference to Figures 4A, 5A, 6, 7, and 9, illustrate a flowchart showing a third embodiment of the object profile image capture method according to the present invention, and a projection coordinate transformation image processing method in which the focal plane corresponds to keystone correction by an image sensor.

[0060] As shown in Figure 18, for ease of understanding, slice images 800, 800a, 800b, 800c, and 800d are all displayed on the image sensing surface 13 in Figure 12. As shown in Figure 18, since the focal length and focus of the camera device 10b are fixed, the object points 92c, 92d, and 92e in the slice images 800a, 800b, 800c, and 800d acquired as the camera device moves correspond to the contour feature point image 820 (image of object point 92c in slice image 800a), the contour feature point image 830 (image of object point 92 in slice image 800a), and the contour feature point image 840 (image of object point 92e in slice image 800a), respectively. As the object points 92c, 92d, and 92e enter the field of view of the camera device, they gradually become sharper as they approach the camera device's focus. As the object points 92c, 92d, and 92e move away from the camera device's field of view, they gradually become blurred as they move away from the camera device's focus. Slice images 800, 800a, 800b, 800c, and 800d display images in which contour feature point images 820, 830, and 840 change from blurry to sharp and then back to blurry, facilitating subsequent evaluation of depth position. It should be noted that, as shown in Figure 18, slice images 800, 800a, 800b, 800c, and 800d are images of the original map, and because the focal plane 15 of the object under test is tilted, the focal plane 15 of the object under test is actually the projected slice images 800*, 800a*, 800b*, 800c*, and 800d* after trapezoidal / rectangular transformation correction (see Figure 19). The present invention describes an image in which, when an image is captured by moving only the lens module according to the third embodiment, the object point on the focal plane corresponds to the image acquisition focus of the image sensor, and the image changes from blurry to sharp and then back to blurry. The image generated when moving the lens module according to the first and second embodiments to capture an image, which changes from blurry to sharp and then back to blurry, will not be described again here.

[0061] As shown in Figures 9 and 5A, when an image sensor 13 is set parallel to the movement path 80 of the first plane 70 in the lens module 11b according to the third embodiment, the point to be measured and the image sensor imaging module 13 maintain a fixed depth distance. Since the sensor deviation angle 17(φ) is equal to the lens deviation angle 18(θ), the state of the image of the object to be measured during the moving imaging process does not change in size magnification in a relative ratio from the in-focus state to the out-of-focus state. Therefore, the trapezoidal correction of each slice before displacement can be extended until S4, saving the computing power of the computer. In this case, the lens module 11b according to the third embodiment, similar to the lens module 11a according to the second embodiment (a telecentric lens is preferred as disclosed in paragraph

[0007] ), finds the focal slice and constructs a 3D object profile by simply shifting the correction amount parallel to the slices at equal distances. If the sensor deflection angle 17(φ) is unequal to the lens deflection angle 18(θ), the lens module 11b according to the third embodiment, like the lens module 11 according to the first embodiment, first performs a geometric deformation matrix transformation of keystone distortion for each slice image (it is necessary to perform step S21). However, as shown in Figure 17, unlike the order in which the lens module 11 according to the first embodiment performs step S21, the lens module 11b according to the third embodiment performs step S21 after step S4 is completed.

[0062] Referring to Figures 19 to 22, the camera device 10 acquires a series of continuous images using the lens module 11b according to the third embodiment (acquiring multiple slice images 800). As shown in Figures 19 and 20, based on the magnification of the camera device 10 and the spacing distance between multiple images, the pixel displacement (offset) of the image regions of adjacent slices is corrected, and a pixel offset stacking process is performed to generate multiple evaluation target images 700 (Figure 13A; the explanation of the related content will not be repeated). At the same time, the focus evaluation calculation process of the Laplacian filter of the depth of focus (DFF) is executed to generate a 3D depth map 710 of the image coordinate system of the object in the image coordinate system 200 (see Figure 21). Finally, as shown in Figure 22, an image space transformation is performed (see step S5) to convert from the 3D depth map 710 of the image coordinate system to a 3D depth map 300 of the world coordinate system (real world), and the accurate 3D depth size (3D data point cloud) of the object under measurement 90 with respect to the first plane 70 (XY plane) is obtained.

[0063] The object profile image capture systems 1 and 1a according to the present invention utilize the characteristic that the movement path 80 of the camera device 10 is not perpendicular to the focal plane 15 on the object side. The camera device according to the present invention only continuously captures images at high speed in the lateral direction (for example, the X-axis or XZ plane). In the prior art, when the object to be measured is larger than the imaging field of view of the camera, the conventional camera device 100a idles without capturing images while moving on the XY mounting movement platform, or while the object to be measured moves to a new field of view of the conventional camera device 100a. This solves the problem of both the imaging utilization rate and detection capability of the conventional camera device being low.

[0064] Although embodiments of the present invention have been described in detail above with reference to the drawings, the specific configuration is not limited to these embodiments, and design modifications and the like are also included within the scope of the gist of the present invention. [Explanation of symbols]

[0065] 1. Object Profile Image Capture System 1a Object Profile Image Capture System 10 Camera device 10a Camera device 10b Camera device 100a Camera device 11 Lens Modules 11a Lens Module 11b Lens Module 111 Lens optical central axis 112 Lens Focal Length 131 Image sensor normal to the imaging plane 13 Image sensor imaging module 300 World Coordinate System 3D Depth Map 151 Object-side focal plane normal 17. Sensor declination angle 18. Lens declination 19 Focal plane declination 20 Mobile device 35. Platform reference surface of the object being measured 35a Platform reference surface of the object to be measured 30. Platform for the object to be measured 30a Platform for the object to be measured 50 Control device 60 Image Processing Modules 61 Evaluation Modules 62 Image Spatial Transformation Module 63 Matrix Transformation Module 70 1st plane 71 1st plane normal 92c Object point 92d object point 92e Object point 91f Top vertex 21. First axis moving guide rail 22 Second axis moving guide rail 80 Travel Paths 80a Travel Route 81a Travel Route 80b Travel Path 80c Travel Route 23. Third axis moving guide rail 90 Object to be measured 90a Measured object 90b Object to be measured 90c Measured object 90d Object to be measured 90e Measured object 90f Measured object 817 Image center reference point 890 Image interval distance 880 slices, thickness 700 images to be evaluated 31 Field of View 33 Unit imaging range T Pixel Offset T' Pixel offset 16a Center point 16b Center point 16c center point 16d center point 16e center point 710 Image Coordinate System 3D Depth Map A Position Pixel Point B Position Pixel Point C Position Pixel Point D Position Pixel Point a Position pixel point b Position pixel point c Position pixel point d Position pixel point 14 Center plane 90 Schheimflug Intersection 15a Focal plane on the object side 15b Focal plane on the object side 15c Focal plane on the object side 15d Focal plane on the object side 15e Focal plane on the object side 0 Focal plane on the object side 1 Focal plane on the object side 2 Focal plane on the object side 3 Focal plane on the object side 4 Focal plane on the object side 800a sliced ​​image 800b slice image 800c slice image 800d slice image 800e slice image 0a Slice image 1a Slice image 2a Slice image 3a Slice image 4a Slice image 800a' ​​Slice image after alignment 800b' Slice image after alignment 800c' Slice image after alignment Slice image after 800d' alignment 800e' Slice image after alignment 800a* Projection slice image after trapezoidal / rectangular transformation correction 800b* Projection slice image after trapezoidal / rectangular transformation correction 800c* Projection slice image after trapezoidal / rectangular transformation correction 800d* Projection slice image after trapezoidal / rectangular transformation correction 800e* Projection slice image after trapezoidal / rectangle transformation correction 820 Contour Feature Point Images 820a Contour feature point image 820b contour feature point image 820c contour feature point image 820d contour feature point image 820e contour feature point image 830 Contour Feature Point Images 830a Contour feature point image 830b Contour Feature Point Image 830c contour feature point image 830d contour feature point image 830e Contour Feature Point Image 840 Contour Feature Point Images 840a Contour feature point image 840b Contour Feature Point Image 840c contour feature point image 840d contour feature point image 840e contour feature point image

Claims

1. An object profile image capture system for analyzing the profile contour of an object under test by evaluating changes in the focus quality of multiple slice images, A camera device comprising a lens module and an image sensor imaging module, wherein the lens module and the image sensor imaging module form the focal plane on the side of the object to be measured, The camera device moves along a movement path relative to the object to be measured, and the movement path is not perpendicular to the focal plane on the object to be measured. An object profile image capture system characterized by comprising: a control device electrically connected to the camera device and the moving device, which controls the movement of the moving device and controls the camera device to acquire the plurality of slice images of the object to be measured when the camera device moves along a movement path relative to the object to be measured.

2. The object profile image capture system according to claim 1, characterized in that the moving device comprises a first axis moving guide rail, a second axis moving guide rail, and a third axis moving guide rail, wherein a first plane is formed by any two of the guide rails of the first axis moving guide rail, the second axis moving guide rail, and the third axis moving guide rail, the camera device moves along the movement path in the first plane with respect to the object to be measured, and a focal plane declination angle is formed by the first plane normal of the first plane and the focal plane normal of the focal plane on the object to be measured side, and the focal plane declination angle is in the range of 0.1° to 60°.

3. The object profile image capture system according to claim 2, characterized in that the lens optical central axis formed by the lens module is perpendicular to the movement path, and the sensor deviation angle formed by the normal of the image sensor imaging plane formed by the image sensor imaging module and the lens optical central axis is in the range of 0.1° to 60°.

4. The object profile image capture system according to claim 2, characterized in that the lens optical central axis formed by the lens module is not perpendicular to the movement path, and the lens declination angle formed by the lens optical central axis and the movement path is in the range of 0.1° to 60°.

5. The object profile image capture system according to claim 4, characterized in that the normal of the image sensor imaging plane is parallel to the optical central axis of the lens.

6. The object profile image capture system according to claim 4, characterized in that the normal of the image sensor imaging plane is not parallel to the optical central axis of the lens, and the sensor deviation angle formed by the normal of the image sensor imaging plane and the optical central axis of the lens is in the range of 0.1° to 60°.

7. The object profile image capture system according to claim 1, characterized in that the imaging frequency of the camera device is in the range of 1 frame / second to 1,000 frames / second.

8. The object profile image capture system according to claim 2, characterized in that the camera device has m unit imaging ranges along the direction of movement, the movement path exhibits a V-shaped path, and when the camera device completes the V-shaped path, the camera device moves from the nth unit imaging range of the m unit imaging ranges to the (n+1)th unit imaging range, where n and m are natural numbers and m > n.

9. The object profile image capture system according to claim 8, characterized in that there is a slice spacing thickness between two adjacent slice images, the V-shaped path has a descending path and an ascending path, and the difference in depth between the endpoint of the descending path and the starting point of the ascending path is 0.5 times the slice spacing thickness.

10. The object profile image capture system according to claim 2, further comprising an image processing module to which a signal is connected to the control device, wherein the camera device moves along the movement path relative to the object to be measured at a magnification ratio, acquires a plurality of slice images of the object to be measured, and has an image spacing distance between adjacent slice images, and the image processing module performs pixel offset stacking on the plurality of slice images based on the magnification ratio, the image spacing distance, and the focal plane deviation to generate a plurality of evaluation target images.

11. The object profile image capture system according to claim 10, characterized in that the image processing module includes an evaluation module that completes a three-dimensional depth map of the image coordinate system by evaluating the plurality of images to be evaluated using a focus evaluation calculation process of a Laplacian filter for depth of focus (DFF).

12. The object profile image capture system according to claim 10, characterized in that the image processing module includes an image space transformation module that transforms the image coordinate system 3D depth map into a world coordinate system 3D depth map based on the focal plane declination and the image spacing distance.

13. The object profile image capture system according to claim 10, wherein the image processing module comprises a matrix transformation module, and if the image sensor imaging module is not parallel to the focal plane on the side of the object to be measured, the matrix transformation module performs a geometric deformation matrix transformation on each of the plurality of slice images before the image processing module performs the pixel offset stacking process on each of the plurality of slice images.

14. The object profile image capture system according to claim 10, wherein the image processing module comprises a matrix transformation module, and if the image sensor imaging module is not parallel to the focal plane on the side of the object to be measured, the matrix transformation module performs a geometric deformation matrix transformation on each of the plurality of slice images after the image coordinate system 3D depth map is completed.

15. A method for capturing object profile images, comprising: moving a camera device relative to an object to be measured using a moving device to acquire multiple slice images of the object to be measured; evaluating the change in focus quality of the multiple slice images to find the profile contour of the object to be measured; the camera device comprising a lens module and an image sensor imaging module, wherein the lens module and the image sensor imaging module form the focal plane on the object side; The camera device moves along a movement path relative to the object to be measured, and the movement path includes steps that are not perpendicular to the focal plane on the object to be measured side. An object profile image capture method characterized by comprising the step of the camera device acquiring a plurality of slice images of the object to be measured as the camera device moves along a movement path relative to the object to be measured.

16. The camera device moves in a first plane along the movement path relative to the object to be measured, and a focal plane deflection angle is formed by the first plane normal of the first plane and the focal plane normal of the focal plane on the object to be measured, and the camera device moves along the movement path relative to the object to be measured with a magnification, and acquires the plurality of slice images of the object to be measured, and there is an image spacing distance between each adjacent slice image, and the method is, The steps include: generating multiple evaluation images by performing pixel offset stacking on the multiple slice images based on the magnification ratio, the image spacing distance, and the focal plane deviation; The object profile image capture method according to claim 15, characterized by comprising the step of completing a three-dimensional depth map of the image coordinate system by evaluating the plurality of images to be evaluated using a Laplacian filter focus evaluation calculation process for depth of focus (DFF).

17. The object profile image capture method according to claim 16, characterized in that the focal plane declination angle is in the range of 0.1° to 60°.

18. The object profile image capture method according to claim 16, further comprising the step of converting the image coordinate system 3D depth map to a world coordinate system 3D depth map based on the focal plane declination angle and the image spacing distance.

19. If the image sensor imaging module is not parallel to the focal plane on the object to be measured, the method performs the pixel offset stacking process on each slice image before proceeding. The object profile image capture method according to claim 16, further comprising the step of performing a geometric transformation matrix transformation on each of the slice images.

20. If the image sensor acquisition module is not parallel to the focal plane on the object to be measured, the method proceeds after the completion of the 3D depth map of the image coordinate system. The object profile image capture method according to claim 16, further comprising the step of performing a geometric transformation matrix transformation on each of the slice images.

Images