Image acquisition control method and device, image acquisition system and readable storage medium

Abstract

Embodiments of the present specification provide an image acquisition control method and device, an image acquisition system and a readable storage medium, wherein the method comprises: when a relative pose between an image acquisition device and an acquisition object continuously changes, acquiring images acquired by the image acquisition device at a plurality of relative poses respectively, and determining imaging change information; based on the imaging change information, evaluating imaging quality of the image acquisition device at each relative pose, and based on an imaging quality evaluation result, determining a relative pose at which the image acquisition device focuses on the acquisition object. By using the above scheme, the image acquisition device can be ensured to focus clearly during image acquisition, and the consistency of the focusing condition during image acquisition can be ensured, thereby effectively improving the imaging quality, imaging stability and universality.

Description

Technical Field

[0001] This specification relates to the field of image acquisition technology, and in particular to an image acquisition control method and device, an image acquisition system and a readable storage medium. Background Technology

[0002] Image acquisition technology can convert information from real space into image data. Then, depending on the application scenario, corresponding image processing operations are performed on the images, and the results are used to meet the needs of the relevant application scenario. For example, in a defect detection scenario, defect detection-related image processing operations can be performed on the images to determine whether the acquired object has defects.

[0003] When acquiring images, the physical properties (such as position and angle) and technological properties (such as structure, size, and material) of the object being acquired can affect the focusing of the image, thereby reducing the image quality and stability and resulting in poor image processing results.

[0004] Therefore, the problem of unsatisfactory focusing during image acquisition needs to be solved by those skilled in the art. Summary of the Invention

[0005] In view of this, embodiments of this specification provide an image acquisition control method and device, an image acquisition system and a readable storage medium, which can ensure that the image acquisition device is in sharp focus during image acquisition and can ensure the consistency of focus during image acquisition, effectively improving imaging quality, imaging stability and universality.

[0006] Specifically, this specification provides an image acquisition control method, including:

[0007] When the relative pose between the image acquisition device and the acquisition object changes continuously, the image acquisition device acquires images at multiple relative poses respectively.

[0008] Based on the acquired multiple images, determine the imaging change information;

[0009] Based on the imaging change information, the imaging quality of the image acquisition device at each of the relative poses is evaluated to obtain the imaging quality evaluation result.

[0010] Based on the imaging quality assessment results, the relative pose of the image acquisition device focusing on the acquisition object is determined.

[0011] Optionally, the continuous change in the relative pose between the image acquisition device and the acquisition object includes:

[0012] Control at least one of the image acquisition device and the acquisition object to perform continuous movement so that the relative pose changes continuously.

[0013] Optionally, determining the imaging change information based on multiple acquired images includes at least one of the following:

[0014] Determine the gradient change information of each of the images;

[0015] A reference image and a corresponding historical image are determined from the plurality of images, and the texture change information of each reference image relative to the historical image is determined.

[0016] Optionally, determining the gradient change information of each of the images includes:

[0017] Extract the gradient features of each image and determine the gradient change information.

[0018] Optionally, extracting the gradient features of each of the images includes:

[0019] Gradient features of each image are extracted based on the multi-point gradient operator of adjacent matrices.

[0020] Optionally, determining the gradient change information of each of the images includes:

[0021] The gradient change information is determined based on a specified region in each of the images.

[0022] Optionally, determining the texture change information of each of the reference images relative to the historical images includes:

[0023] Each of the aforementioned reference images and each of the aforementioned historical images are used as images to be processed, in order to extract the contour features of the images to be processed;

[0024] Based on the extracted contour features, the texture change information is determined.

[0025] Optionally, extracting the contour features of the image to be processed includes:

[0026] According to a specified direction, multiple pixel groups are selected around a specified pixel in the image to be processed, and the gradient of each pixel group of the specified pixel is determined. Based on the gradient of each pixel group of the specified pixel, the contour features of the image to be processed are determined.

[0027] Optionally, the step of selecting multiple pixel groups around a specified pixel in the image to be processed according to a specified direction, determining the gradient of each pixel group of the specified pixel, and determining the contour features of the image to be processed based on the gradients of each pixel group of the specified pixel includes:

[0028] According to the first direction, multiple first pixel groups are selected around a specified pixel in the image to be processed, and after calculating the gradient of each first pixel group corresponding to the specified pixel, the contour features of the image to be processed in the first direction are determined based on the gradient of each first pixel group of the specified pixel.

[0029] In accordance with the second direction, multiple second pixel groups are selected around a specified pixel in the image to be processed, and after calculating the gradient of each second pixel group corresponding to the specified pixel, the contour features of the image to be processed in the second direction are determined based on the gradient of each second pixel group of the specified pixel.

[0030] Based on the contour features of the image to be processed in the first direction and the contour features in the second direction, the contour features of the image to be processed are determined.

[0031] Optionally, before performing a weighted calculation on the gradients of each pixel group of the specified pixel point, the method further includes:

[0032] Based on the gradient change rate among the multiple pixel groups of the specified pixel point, a contour weight is assigned to the gradient of each pixel group.

[0033] Optionally, determining the texture change information of each of the reference images relative to the historical images includes:

[0034] Texture change information is determined based on specified regions in each of the reference images and specified regions in each of the historical images.

[0035] Optionally, the step of evaluating the imaging quality of the image acquisition device at each of the relative poses based on the imaging change information to obtain an imaging quality evaluation result includes:

[0036] The gradient change information and texture change information corresponding to each relative pose are weighted and calculated to obtain the corresponding imaging quality evaluation information.

[0037] Based on the relative poses and corresponding imaging quality assessment information, the imaging quality assessment result is determined.

[0038] Optionally, determining imaging change information based on multiple acquired images includes:

[0039] Based on multiple acquired images, imaging change information is determined multiple times to obtain multiple imaging change information.

[0040] The process of evaluating the imaging quality of the image acquisition device at each relative pose based on the imaging change information, and obtaining imaging quality evaluation results, includes:

[0041] Based on the imaging change information, the imaging quality of the image acquisition device at each of the relative poses is evaluated to obtain the corresponding candidate quality evaluation results.

[0042] Match multiple candidate quality assessment results;

[0043] After the matching result meets the confidence criteria, the imaging quality assessment result is determined based on multiple candidate quality assessment results.

[0044] This specification also provides an image acquisition control device, connected to the image acquisition device, including:

[0045] An image acquisition unit is adapted to acquire images acquired by the image acquisition device at multiple relative poses when the relative pose between the image acquisition device and the acquisition object changes continuously.

[0046] The information acquisition unit is suitable for determining imaging change information based on multiple acquired images;

[0047] The evaluation unit is adapted to evaluate the imaging quality of the image acquisition device at each of the relative poses based on the imaging change information, obtain the imaging quality evaluation result, and determine the relative pose of the image acquisition device focusing on the acquisition object based on the imaging quality evaluation result.

[0048] This specification also provides an image acquisition and control device, including a memory and a processor. The memory stores computer instructions that can be executed on the processor. When the processor executes the computer instructions, it performs the steps of the method described in any of the above embodiments.

[0049] This specification also provides an image acquisition system, including:

[0050] Loading equipment, suitable for loading the objects to be collected;

[0051] An image acquisition device, suitable for acquiring images of the object to be acquired;

[0052] An image acquisition control device is connected to an image acquisition device; it is adapted to acquire images acquired by the image acquisition device at multiple relative poses when the relative pose between the image acquisition device and the acquisition object changes continuously; based on the acquired multiple images, it determines imaging change information; based on the imaging change information, it evaluates the imaging quality of the image acquisition device at each of the relative poses to obtain an imaging quality evaluation result; based on the imaging quality evaluation result, it determines the relative pose of the image acquisition device focusing on the acquisition object.

[0053] This specification also provides a computer-readable storage medium storing computer instructions that, when executed, perform the steps of the methods described in any of the above embodiments.

[0054] Using the image acquisition control method provided in the embodiments of this specification, when the relative pose between the image acquisition device and the acquisition object changes continuously, images acquired by the image acquisition device at multiple relative poses are obtained to determine imaging change information; then, based on the imaging change information, the imaging quality of the image acquisition device at each relative pose is evaluated, and based on the imaging quality evaluation results, the relative pose of the image acquisition device focusing on the acquisition object is determined. As can be seen from the above, since multiple images are acquired by the image acquisition device during continuous changes in relative pose, the influence of the surrounding environment on image acquisition can be reduced, improving the timeliness of image acquisition. By evaluating the imaging quality of the image acquisition device through imaging change information, the obtained imaging quality evaluation results can accurately reflect the true focusing situation of the image acquisition device. Based on the imaging quality evaluation results, the changing trend of the true focusing situation of the image acquisition device can be obtained. Therefore, according to the actual situation, the relative pose that meets the expected focusing effect can be adaptively determined. Thus, the image acquisition control method provided in this specification can ensure that the image acquisition device has clear focus during image acquisition and can ensure the consistency of the focusing situation during image acquisition, effectively improving imaging quality and imaging stability, and thereby improving the accuracy and reliability of image processing results. Furthermore, the image acquisition control method provided in this specification is applicable to application scenarios where there are at least one of the following: variable acquisition objects and batch image acquisition, thus having higher versatility. Attached Figure Description

[0055] To more clearly illustrate the technical solutions of the embodiments of this specification, the drawings used in the description of the embodiments of this specification or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0056] Figure 1 This is a flowchart of an image acquisition control method provided in an embodiment of this specification.

[0057] Figure 2 This is a schematic diagram of pixel group selection provided in an embodiment of this specification.

[0058] Figure 3a This is a schematic diagram of an image of a circuit board acquired by an image acquisition device in the current pose, as provided in an embodiment of this specification.

[0059] Figure 3b for Figure 3a A schematic diagram of the parabola between the relevant relative pose and image quality assessment information.

[0060] Figure 3c For image acquisition devices Figure 3b The image of the circuit board acquired under the relative pose corresponding to the top of the parabola shown.

[0061] Figure 4 A structural block diagram of an image acquisition and control device provided in the embodiments of this specification.

[0062] Figure 5 This is a structural block diagram of another image acquisition and control device provided in the embodiments of this specification.

[0063] Figure 6 This is a structural block diagram of an image acquisition system provided in an embodiment of this specification. Detailed Implementation

[0064] As described in the background section, the physical attributes (such as position, angle, etc.) and technological attributes (such as structure, color, texture, material, etc.) of the object being acquired can affect the focusing during image acquisition, thereby reducing imaging quality and imaging stability, resulting in poor image processing results.

[0065] For example, for the same object A1, if the position or angle of object A1 shifts, the focus will also change, potentially resulting in a blurry image and reduced image quality. Furthermore, if multiple images of object A1 are acquired, multiple images with different imaging effects may be obtained. Therefore, if image processing operations are performed, different processing results may be obtained depending on the timing of image acquisition, thus reducing the accuracy and reliability of the image processing results.

[0066] For example, when the acquired object A2 is replaced with acquired object A3, the focus will change because the surfaces of acquired object A2 and acquired object A3 have different details. This may cause the image of acquired object A3 to become blurry, thus reducing image quality. Therefore, when performing image processing operations on the image of acquired object A3, the error rate of the image processing results increases, thereby reducing the accuracy and reliability of the image processing results.

[0067] As can be seen from the above, even for the same object, changes in its position, angle, etc., will alter the focusing, which can easily affect the image quality and stability, leading to poor image processing results. For multiple objects, the focusing is even more complex and variable, which can also easily affect the image quality and stability, resulting in poor image processing results.

[0068] To address the issue of unsatisfactory focusing during image acquisition, this specification provides an image acquisition control method. When the relative pose between the image acquisition device and the target object continuously changes, images acquired by the image acquisition device at multiple relative poses are obtained to determine imaging change information. Based on this imaging change information, the imaging quality of the image acquisition device at each relative pose is evaluated. Based on the imaging quality evaluation results, the relative pose of the image acquisition device focused on the target object is determined. This ensures clear focusing of the image acquisition device during image acquisition and guarantees consistency in focusing, effectively improving imaging quality, imaging stability, and versatility.

[0069] To enable those skilled in the art to better understand and implement the concept, solution, and advantages of the present invention, the technical solution is described in detail below with reference to the accompanying drawings and specific embodiments.

[0070] Reference Figure 1 This is a flowchart illustrating an image acquisition control method provided in an embodiment of this specification. In this embodiment, the image acquisition control method may include:

[0071] S01, when the relative pose between the image acquisition device and the acquisition object changes continuously, the image acquisition device acquires images at multiple relative poses respectively.

[0072] The image can be at least one of a color image and a grayscale image. The relative pose can include at least one of the following: the relative position between the image acquisition device and the acquisition object; the relative angle between the image acquisition device and the acquisition object.

[0073] In specific implementations, the relative pose can be described according to a coordinate system. For example, after establishing an XYZ three-dimensional coordinate system, the relative pose between the image acquisition device and the acquisition object can include at least one of the following: the relative position between the image acquisition device and the acquisition object along the X-axis; the relative position between the image acquisition device and the acquisition object along the Y-axis; the relative position between the image acquisition device and the acquisition object along the Z-axis; the relative angle of rotation between the image acquisition device and the acquisition object about the X-axis; the relative angle of rotation between the image acquisition device and the acquisition object about the Y-axis; and the relative angle of rotation between the image acquisition device and the acquisition object about the Z-axis.

[0074] S02, based on the acquired multiple images, determine the imaging change information.

[0075] In specific implementations, when the image acquisition device acquires images at multiple relative poses, it can simultaneously record the relative poses corresponding to each image. Therefore, a correspondence exists between the images and the relative poses, and the imaging change information obtained based on the images is suitable for characterizing the imaging changes of the image acquisition device when the relative poses change continuously. The imaging change information can be represented by corresponding computer-recognizable characters. This specification does not limit the representation of the imaging change information in the embodiments.

[0076] S03, based on the imaging change information, the imaging quality of the image acquisition device at each of the relative poses is evaluated to obtain the imaging quality evaluation result.

[0077] In specific implementations, the imaging quality assessment results can be used to characterize the imaging quality of the image acquisition device at various relative poses when the relative pose changes continuously. The imaging quality assessment results can be represented by corresponding computer-recognizable characters. This specification does not limit the specific representation of the imaging quality assessment results in the embodiments.

[0078] In specific implementation, the imaging quality assessment result may include: imaging quality assessment information corresponding to each relative pose, which can represent the imaging quality status through text or numerical values. For example, the text "pass" can be used to represent that the imaging quality status at the relative pose is passable; or the text "fail" can be used to represent that the imaging quality status at the relative pose is failable; or, for another example, the imaging quality status at the relative pose can be represented by numerical values.

[0079] In practice, imaging quality assessments can be conducted based on specific imaging quality requirements. For example, if there is a requirement for sharpness, an imaging quality assessment in terms of sharpness can be performed; correspondingly, the computer-recognizable characters included in the imaging quality assessment results can characterize the imaging quality assessment results in terms of sharpness.

[0080] S04, based on the imaging quality assessment results, determine the relative pose of the image acquisition device focusing on the acquisition object.

[0081] In practical implementation, as the relative pose between the image acquisition device and the acquired object continuously changes, there will be a relative pose where the image acquisition device focuses on the acquired object. At this relative pose, the image acquisition device can be considered to be in sharp focus, while at other relative poses, the image acquisition device is not in focus on the acquired object, i.e., the image acquisition device is in blurry focus. Based on the imaging quality assessment results, after determining the imaging quality assessment information that meets the preset selection conditions, the corresponding relative pose can be selected as the relative pose for the image acquisition device to focus on the acquired object.

[0082] Based on this, when the image acquisition device is controlled to acquire images at the selected relative pose, the image acquisition device is clearly focused and can obtain images with good imaging quality.

[0083] Furthermore, since the image acquisition device can ensure clear focus during image acquisition, it can not only improve imaging quality but also improve the consistency of focusing when there are at least one of the following situations in the application scenario: batch image acquisition and multiple acquisition objects.

[0084] For example, for multiple acquisition objects with the same process attributes, even if the physical attributes of these acquisition objects change (such as changes in position or angle), after determining the relative pose based on the imaging quality assessment results, the image acquisition device can meet the unified focusing requirements when facing multiple acquisition objects with different physical attributes, so that the focusing situation when acquiring images of each acquisition object remains consistent, thereby ensuring both imaging quality and imaging stability.

[0085] For multiple acquisition objects with the same physical properties but different process properties, and for multiple acquisition objects with different physical and process properties, after determining the relative pose based on the imaging quality assessment results, the image acquisition device can meet the unified focusing requirements when facing these acquisition objects, so that the focusing situation when acquiring images of each acquisition object remains consistent, thereby ensuring both imaging quality and imaging stability.

[0086] In practical implementation, selection conditions can be set according to the representation method of imaging quality assessment information, thereby selecting from multiple relative poses. It is understood that the embodiments in this specification do not limit the specific content of the selection conditions.

[0087] For example, when the imaging quality assessment information represents the imaging quality status using the words "pass" and "fail," the selection condition can be set as follows: select the relative pose corresponding to the first imaging quality assessment information with the word "pass." As another example, when the imaging quality assessment information represents the imaging quality status using numerical values, the selection condition can be set as follows: select the relative pose corresponding to the highest numerical value.

[0088] As can be seen from the above, since multiple images are acquired by the image acquisition device during the continuous change of relative pose, the influence of the surrounding environment on image acquisition can be reduced, and the timeliness of image acquisition can be improved. By evaluating the imaging quality of the image acquisition device through imaging change information, the obtained imaging quality evaluation result can accurately reflect the true focusing situation of the image acquisition device. Based on the imaging quality evaluation result, the changing trend of the true focusing situation of the image acquisition device can be obtained. Therefore, the relative pose that meets the expected focusing effect can be adaptively determined according to the actual situation. Thus, the image acquisition control method provided in this specification can ensure that the image acquisition device is clearly focused during image acquisition and can ensure the consistency of the focusing situation during image acquisition, effectively improving imaging quality and imaging stability, and thereby improving the accuracy and reliability of image processing results.

[0089] Furthermore, the image acquisition control method provided in the embodiments of this specification is applicable to application scenarios where there are at least one of the following: variable acquisition objects and batch image acquisition, thus having greater versatility.

[0090] In practice, when determining imaging change information and evaluating imaging quality, the relative pose between the image acquisition device and the acquisition object can be continuously changed, and the image acquisition device acquires images.

[0091] Since the image acquisition control method provided in the embodiments of this specification needs to acquire multiple images, and the method steps may be executed synchronously, for ease of description and understanding, they can be sorted according to the acquisition order, and related processing steps (such as executing...) can be performed in sequence. Figure 1 In step S2 or S3 (or similar steps), a feature can be determined as a reference, i.e., a reference feature. Features preceding the reference feature can be called historical features. For example, an image can be determined as a reference, i.e., a reference image. Images acquired before the reference image can be called historical images.

[0092] Understandably, the selection method for baseline and historical features should be determined based on actual processing efficiency. For example, stored images can be selected sequentially according to their order, or the most recently acquired image can be used as the baseline image.

[0093] Furthermore, there can be one or more relative poses preceding the reference image, meaning there can be one or more historical images. For the first relative pose (i.e., the initial relative pose), related method steps (such as...) may not be performed. Figure 1 Alternatively, step S2 or step S3 can be used to preset imaging change information and imaging quality evaluation information for the initial relative pose.

[0094] It should be noted that the above examples are for illustrative purposes only. In practical applications, the images can be replaced with other features with temporal characteristics (such as relative pose) in the embodiments of this specification. This specification does not impose any specific restrictions on this.

[0095] In practice, during the continuous change of the relative pose between the image acquisition device and the acquisition object, there will be times when the image acquisition device focuses on the relative pose of the acquisition object, at which point the imaging quality is optimal. However, at other relative poses, the imaging quality is not optimal.

[0096] Based on this, the optimal imaging quality assessment information can be determined by comparing the imaging quality assessment information corresponding to the reference relative pose with the imaging quality assessment information corresponding to the historical relative pose. Then, the relative pose corresponding to the optimal imaging quality assessment information is used as the relative pose for the image acquisition device to focus on the acquisition object.

[0097] In an optional example, in order to obtain a more accurate relative pose of the image acquisition device focusing on the acquisition object, curve fitting can be performed based on the imaging quality assessment information corresponding to each relative pose included in the imaging quality assessment result to obtain a parabola about the relative pose and the imaging quality assessment information, and the relative pose corresponding to the top of the parabola can be taken as the relative pose of the image acquisition device focusing on the acquisition object.

[0098] In specific implementation, depending on the specific situation and needs, at least one of the image acquisition device and the acquisition object can be controlled to move continuously so that the relative pose changes continuously. During the process of the continuous change of the relative pose, the image acquisition device can be controlled to acquire multiple images of the acquisition object to obtain multiple images.

[0099] When controlling one of the image acquisition device and the acquisition object to move continuously, the relative pose can be represented by the position and / or angle of the moving subject (i.e., one of the image acquisition device and the acquisition object).

[0100] In practical implementation, the movement speed of at least one of the image acquisition device and the acquisition object can be controlled according to the specific application scenario and requirements. For example, at least one of the image acquisition device and the acquisition object can be controlled to move at a constant speed continuously. Alternatively, at least one of the image acquisition device and the acquisition object can be controlled to move at a variable speed continuously.

[0101] In practical implementation, the continuous movement of at least one of the image acquisition device and the acquisition object can be controlled according to specific application scenarios and requirements. For example, the image acquisition device and at least one of the acquisition object can be controlled to perform continuous movement in a single direction. Another example is that the image acquisition device and at least one of the acquisition object can be controlled to perform reciprocating continuous movement.

[0102] In practical implementation, since the imaging quality assessment information corresponding to multiple relative poses can reflect the changing trend of the actual focus of the image acquisition device, the direction of change of the relative pose can be adjusted by comparing the imaging quality assessment information corresponding to the reference relative pose and the imaging quality assessment information corresponding to the historical relative pose. For example, when the relative pose continuously increases, by comparing the imaging quality assessment information corresponding to the reference relative pose and the imaging quality assessment information corresponding to the historical relative pose, it can be determined that the actual focus of the image acquisition device is becoming increasingly blurred, and the relative pose can be continuously decreased. This improves processing efficiency and allows for a faster determination of the relative pose of the image acquisition device focused on the acquisition object.

[0103] In specific implementations, the imaging change information can be determined based on information related to the visualization layer contained in multiple images. For example, the process of determining imaging change information based on multiple acquired images may include at least one of the following:

[0104] 1) Determine the gradient change information of each of the images; wherein the gradient change information is suitable for characterizing the change in grayscale of the image itself.

[0105] 2) Determine a reference image and a corresponding historical image from the plurality of images respectively, and determine the texture change information of each reference image relative to the historical image; wherein the texture change information is suitable for characterizing the texture change situation.

[0106] It is understandable that for the first image acquired by the image acquisition device during the relative pose change process, there is actually no corresponding historical image when it is used as the reference image. In this case, corresponding texture change information can be preset for the first image through prior conditions, laboratory data, etc. (e.g., setting the texture change information of the first image to 0). Alternatively, considering that the probability of the first image being the clearest image is relatively small, it can be set not to select the first image as the reference image. The embodiments in this specification do not impose specific limitations on this.

[0107] In practical implementation, during the process of determining gradient change information, the gradient features of each image can be extracted first, and the gradient change information can be determined. The gradient features are suitable for characterizing the grayscale changes of the image itself.

[0108] It should be noted that the specific method for extracting gradient features can be determined based on the specific application scenario and requirements.

[0109] In specific implementations, the process of extracting gradient features from each image may include: extracting gradient features from each image based on a multi-point gradient operator of adjacent matrices. The multi-point gradient operator of adjacent matrices may include one or more combinations of the following: Roberts operator, Prewitt operator (a first-order differential operator), Sobel operator, Laplace operator, and custom operators.

[0110] Depending on the adjacent matrix multi-point gradient operator used in the specific implementation, a variety of gradient features can be obtained. When multiple gradient features are obtained, one or more gradient features can be calculated according to the specific application scenario to obtain the gradient change information.

[0111] Taking the Sobel operator as an example, a horizontal filter template and a vertical filter template are set. The horizontal filter template moves on the image and performs neighborhood convolution with the image to obtain multiple gradients in the horizontal direction of the image. Correspondingly, the vertical filter template also moves on the image and performs neighborhood convolution with the image to obtain multiple gradients in the vertical direction of the image. There is a corresponding relationship between the gradients in the horizontal direction of the image and the gradients in the vertical direction of the image, depending on the moving direction and the moving step size.

[0112] Then, based on the gradients in the horizontal and vertical directions of the image, a modulo operation is performed to obtain the gradient magnitude of the image, and an arctangent operation is performed to obtain the gradient direction of the image.

[0113] As shown above, the Sobel operator can yield four gradient features: the horizontal gradient, the vertical gradient, the gradient magnitude, and the gradient direction. In applications where the surface features of the acquired object are relatively simple (such as line measurement), only the horizontal and vertical gradients can be calculated to obtain the gradient change information. This reduces the computational load of gradient change information and ensures the accuracy of the image acquisition control method provided in this specification. In applications where the surface features of the acquired object are more complex (such as device detection), the horizontal, vertical, gradient magnitude, and gradient direction can be calculated to obtain the gradient change information. This improves the accuracy and reliability of the gradient change information.

[0114] In practice, in order to effectively reduce the amount of computation while retaining more useful information, the gradient change information of each image can be determined based on a specified region in each image during the process of determining the gradient change information of each image.

[0115] The designated area can be the area containing the most useful information (or containing relatively concentrated useful information) obtained through a preset target box.

[0116] In practice, the continuous change in the relative pose between the image acquisition device and the object being acquired will affect the focusing situation. As a result, the image acquisition device will acquire both blurry and clear images. The clear image has more detailed information than the blurry image. For example, more accurate and richer contours can be identified from the clear image.

[0117] The combination of contours in an image can be considered as a texture of the image. Based on this, in the process of determining texture change information, each of the reference images and each of the historical images can be used as images to be processed to extract the contour features of the images to be processed. Then, based on the extracted contour features, the texture change information can be determined.

[0118] The contour features are suitable for characterizing the distribution of contours in an image. It should be noted that the specific method for extracting the contour features can be determined based on the specific application scenario and requirements.

[0119] In one optional example, image WN is determined as the reference image among multiple images, and image WN-1, which precedes and is adjacent to the reference image WN, is designated as the historical image. Then, the reference image WN and the historical image WN-1 are used as images to be processed, and contour features are extracted to obtain the contour features of the reference image WN and the historical image WN-1. Then, based on the extracted contour features of the reference image WN and the historical image WN-1, the texture change information of the reference image WN relative to the historical image WN-1 can be determined.

[0120] In a specific implementation, the extraction of the contour features of the image to be processed may include: selecting multiple pixel groups around a specified pixel point in the image to be processed according to a specified direction, determining the gradient of each pixel group of the specified pixel point, and determining the contour features of the image to be processed based on the gradient of each pixel group of the specified pixel point.

[0121] It is understandable that, depending on the specific circumstances, one or more directions can be specified. Furthermore, to facilitate direction selection, at least one of the horizontal and vertical directions of the image can be selected as the specified direction.

[0122] In an optional example, multiple first pixel groups are selected around a specified pixel in the image to be processed in a first direction, and after calculating the gradient of each first pixel group corresponding to the specified pixel, the contour features of the image to be processed in the first direction are determined based on the gradient of each first pixel group of the specified pixel.

[0123] In accordance with the second direction, multiple second pixel groups are selected around a specified pixel in the image to be processed. After calculating the gradient of each second pixel group corresponding to the specified pixel, the contour features of the image to be processed in the second direction are determined based on the gradient of each second pixel group of the specified pixel.

[0124] Then, based on the contour features of the image to be processed in the first direction and the contour features in the second direction, the contour features of the image to be processed are determined.

[0125] In practice, to facilitate the selection of pixel groups, multiple regions can be divided around the specified pixel point according to a specified direction, and pixels can be selected in each region to obtain the corresponding pixel group.

[0126] In an optional example, such as Figure 2 The diagram shown is a schematic representation of pixel group selection provided in an embodiment of this specification. Figure 2In the image P1 to be processed, there are m rows and n columns of pixels. Image P1 can be viewed as a two-dimensional function F(x, y), and each pixel in image P1 can be represented by the two-dimensional function Z = F(x, y). For example, Z... 1,1 =F(x1, y1) can be represented as the pixel in the first row and first column.

[0127] In the image P1 to be processed, the pixel Z in the i-th row and j-th column can be... i,j As a specified pixel point, where i and j are positive integers, and i is less than m, j is less than n. Following a specified direction (horizontal in this example), at the specified pixel point Z... i,j The area around it is divided into several zones (in Figure 2 (Distinguished by different fill lines), namely Region 1, Region 2, and Region 3. Region 1 may include pixel Z. i-1,j-1 Z i-1,j and Z i-1,j+1 Region 2 can include pixel Z. i,j-1 Z i,j and Z i,j+1 Region 3 can include pixel Z. i+1,j-1 Z i+1,j and Z i+1,j+1 .

[0128] Select pixel Z in region 1 i-1,j-1 and Z i-1,j+1 This forms pixel group 1. Select Z in region 2. i,j-1 and Z i,j+1 This forms pixel group 2. Z is selected within region 3. i+1,j-1 and Z i+1,j+1 This forms pixel group 3.

[0129] In a specific implementation, after determining the gradient of each pixel group of the specified pixel point, a weighted operation can be performed on the gradient of each pixel group of the specified pixel point to determine the contour features of the image to be processed.

[0130] In an optional example, continue to refer to Figure 2 After obtaining the specified pixel Z i,j After calculating pixel groups 1 to 3, the gradients of pixel groups 1 to 3 are calculated respectively, and the gradients of pixel groups 1 to 3 are weighted to obtain the Z value of the specified pixel point. i,j The corresponding contour features.

[0131] For ease of understanding, the above content can be expressed by the following formula:

[0132] T i,j =k1×(Z i,j+1 -Zi,j-1 )+k2×(Z i-1,j+1 -Z i-1,j-1 )+k3×(Z i+1,j+1 -Z i+1,j-1 );

[0133] Among them, T i,j Indicates the specified pixel Z i,j The corresponding contour features, (Z i,j+1 -Z i,j-1 (Z) represents the gradient of pixel group 2. i-1,j+1 -Z i-1,j-1 (Z) represents the gradient of pixel group 1. i+1,j+1 -Z i+1,j-1 ) represents the gradient of pixel group 3. k1, k2 and k3 are the contour weights of pixel group 2, pixel group 1 and pixel group 3, respectively. The contour weights are suitable for characterizing the importance of the corresponding pixel group in the process of obtaining the contour features of a specified pixel.

[0134] Based on the contour features corresponding to selected pixels in the image P1 to be processed, the contour features of the image P1 to be processed can be obtained. Specifically, the contour features of the image P1 to be processed... It can be:

[0135]

[0136] Among them, T 2,2 To T m-1,n-1 These are the contour features of the pixels in the 2nd row and 2nd column, up to the contour features of the pixels in the (m-1)th row and (n-1)th column.

[0137] In practice, in order to obtain the contour features of edge pixels (such as pixels in the first row, first column, last row, and last column of the image to be processed), pixel filling can be performed outside the edge pixels (i.e., around the image to be processed) before selecting multiple pixel groups, thereby expanding the image to facilitate obtaining the contour features of the edge pixels.

[0138] In an optional example, in conjunction with the reference Figure 2 Before selecting multiple pixel groups, pixel filling can be performed on the top and left sides of the image P1 to be processed, thereby expanding the m*n size image P1 to an image P1' of size (m+2)*(n+2). This, in turn, enhances the contour features of the image P1'. It can be:

[0139]

[0140] In practice, to facilitate pixel filling, pixel filling can be performed by copying the corresponding edge pixels.

[0141] In practice, to improve the accuracy and flexibility of the contour feature acquisition process, contour weights can be assigned to the gradient of each pixel group based on the gradient change rate between multiple pixel groups of the specified pixel point.

[0142] The contour change rate can be the ratio of the gradient difference between a pixel group and its preceding pixel group to the gradient of that pixel group. For example, refer to... Figure 2 In the example described, the gradient change rate between pixel group 2 and pixel group 1 can be: [(Z i,j+1 -Z i,j-1 )-(Z i-1,j+1 -Z i-1,j-1 )] / (Z i,j+1 -Z i,j-1 ).

[0143] It is understood that, for ease of description and understanding, the above example only schematically illustrates the case of obtaining the contour features of the image to be processed according to a specified direction. However, in practical applications, there may be multiple specified directions. Furthermore, based on the above, the method for obtaining contour features when there are multiple specified directions can be deduced. This specification does not impose any specific limitations on this.

[0144] It is understood that, for ease of description and understanding, the above example only schematically shows the region of adjacent pixels around a specified pixel. However, in practical applications, a larger region can be selected, and more pixels can be selected as pixel groups within the region. Furthermore, based on the above, the method for obtaining pixel group gradients can be derived, and this specification does not impose specific limitations on this.

[0145] In specific implementation, based on the contour features extracted from the image to be processed, contour regions in the image to be processed are determined, and based on the contour regions in the image to be processed, region gradient features are extracted to obtain the texture information of the image to be processed. Then, the texture information obtained by using the reference image as the image to be processed and the texture information obtained by using the historical image as the image to be processed are compared to determine the texture change information. The extraction process of the region gradient features can refer to the gradient extraction process of the image described above, and will not be repeated here. Optionally, the granularity of the extraction of the region gradient features can be smaller than the granularity of the extraction of the contour features.

[0146] In one optional example, contour regions in a reference image can be determined based on contour features extracted from the reference image. Then, gradient features of these contour regions are extracted and used as texture information for the reference image. Similarly, contour regions in historical images can be determined based on contour features extracted from historical images. Gradient features of these historical images are extracted and used as texture information for the historical images. Finally, texture change information can be determined based on the texture information of the reference image and the historical image.

[0147] By employing the above scheme, the contour regions in the image can be determined through the extracted contour features. Since there is relatively obvious edge information in the contour regions, the probability of texture information is relatively high. Therefore, texture information can be accurately judged and obtained, thus improving the accuracy of texture information. Then, by extracting regional gradient features from the contour regions, the contour regions can be analyzed more finely, thereby improving the accuracy of texture information determination.

[0148] In practice, in order to effectively reduce image size while retaining more useful information, the texture change information can be determined based on a specified region in each of the reference images and a specified region in each of the historical images.

[0149] The designated area can be the area containing the most useful information (or containing relatively concentrated useful information) obtained through a preset target box.

[0150] In specific implementations, both the gradient change information and texture change information can be represented numerically. Based on this, after determining the gradient change information and texture change information corresponding to each relative pose, a weighted calculation can be performed on the gradient change information and texture change information corresponding to each relative pose to obtain the corresponding imaging quality assessment information. Based on each relative pose and the corresponding imaging quality assessment information, the imaging quality assessment result is determined. The weights of the gradient change information and the texture change information can be set according to specific requirements.

[0151] Therefore, by using weighted calculations, the importance of the gradient change information and the texture change information in the imaging quality assessment results can be flexibly controlled, thereby obtaining imaging quality assessment results with more reference value.

[0152] In practice, to improve the reliability of the imaging quality assessment process, imaging change information of the same set of images is acquired multiple times, and the imaging quality assessment results are matched to determine whether there are any anomalies (such as garbled characters, data packet loss, etc.) during the imaging quality assessment process. It should be noted that, for ease of description and understanding, the imaging quality assessment results of the same set of images acquired multiple times are referred to as candidate quality assessment results.

[0153] Specifically, based on multiple acquired images, imaging change information is determined multiple times to obtain multiple imaging change information; and based on each of the imaging change information, the imaging quality of the image acquisition device at each of the relative poses is evaluated to obtain corresponding candidate quality evaluation results; then, the multiple candidate quality evaluation results are matched, and after the matching results meet the confidence conditions, the imaging quality evaluation result is determined based on the multiple candidate quality evaluation results.

[0154] In practical implementation, a matching process for multiple candidate quality assessment results can be set according to specific needs and application scenarios. For example, at least one of the mean, variance, and standard deviation of multiple candidate quality assessment results can be used as the matching result. Alternatively, any two candidate quality assessment results can be matched; if the two results are identical or within the allowable error range, they are considered matched, and the number of matches can be used as the matching result. This specification does not impose specific limitations on this aspect in the embodiments.

[0155] Furthermore, corresponding credibility conditions can be set according to the specific matching process to determine whether the credibility conditions are met. For example, in the matching process of multiple candidate quality assessment results, when at least one of the mean, variance, and standard deviation of multiple candidate quality assessment results is used as the matching result, the credibility condition can be: if the matching result exceeds a credibility threshold, the matching result meets the credibility condition; otherwise, the matching result does not meet the credibility condition. As another example, in the matching process of multiple candidate quality assessment results, when the number of mutually matching candidate quality assessment results is used as the matching result, the credibility condition can be: if the matching result exceeds a matching quantity threshold, the matching result meets the credibility condition; otherwise, the matching result does not meet the credibility condition.

[0156] It should be noted that, depending on the actual application scenario, "exceeding" can be understood as "greater than" or "greater than or equal to". The embodiments in this specification do not impose specific limitations in this regard.

[0157] In practice, to improve the accuracy of imaging change information, each image can be preprocessed before determining the imaging change information, so that the preprocessed image can be used to determine the imaging change information.

[0158] The preprocessing includes at least one of the following: converting the image to grayscale; and performing target region filtering on the image.

[0159] In specific implementation, the target region filtering process for the image may include: filtering out the foreground regions in each image using a foreground segmentation algorithm. In an optional example, the foreground segmentation method may include one or more of the following: the big rule algorithm, a threshold-based histogram segmentation algorithm, a topology-based watershed algorithm, a clustering-based segmentation algorithm, and a mathematical morphology-based segmentation algorithm.

[0160] In practical implementation, the image acquisition control method provided in the embodiments of this specification can be widely applied to various application scenarios involving image acquisition, such as shooting scenarios, defect detection scenarios, key size detection scenarios, key point detection scenarios, target recognition scenarios, etc. Accordingly, depending on the specific application scenario, the images in the embodiments of this specification can be acquired from objects such as people, objects, and environments. This specification does not impose specific limitations on the application scenarios or acquisition objects.

[0161] For example, in critical dimension inspection scenarios on circuit boards, such as Figure 3a The image shown is a schematic diagram of an image of a circuit board captured by an image acquisition device in the current pose. (Reference) Figure 3a It can be seen that the circuit board image captured by the image acquisition device in the current pose is relatively blurry.

[0162] As the relative pose between the image acquisition device and the target object continuously changes, images of the circuit board are acquired by the image acquisition device at multiple relative poses. Then, based on the acquired images of the circuit board, imaging change information is determined, and based on this imaging change information, the imaging quality of the image acquisition device at each relative pose is evaluated to obtain an imaging quality evaluation result. Furthermore, based on the imaging quality evaluation information corresponding to each relative pose included in the imaging quality evaluation result, curve fitting is performed to obtain a parabola about the relative pose and the imaging quality evaluation information, such as... Figure 3b As shown, the relative pose WZ corresponding to the apex of the parabola is taken as the relative pose of the image acquisition device focusing on the acquisition object. Figure 3c This refers to the image of the circuit board captured by the image acquisition device in the relative pose (i.e., relative pose WZ) corresponding to the apex of the parabola. (Comparison) Figure 3a and Figure 3cAs can be seen, the image acquisition control method provided in the embodiments of this specification can enable the image acquisition device to focus on the acquisition object, thereby ensuring that the image acquisition device is clearly focused during image acquisition and can ensure the consistency of the focusing situation during image acquisition, effectively improving the imaging quality and imaging stability, and thus improving the accuracy and reliability of the image processing results.

[0163] In practical implementation, the imaging quality of an image acquisition device is related not only to the relative pose between the image acquisition device and the acquired object, but also to optical parameters. These optical parameters may include: the resolution parameters of the image acquisition device, the objective lens parameters of the image acquisition device, the type of light source in the environment where the image acquisition device is located (e.g., bright field light source, dark field light source, etc.), and the lighting conditions of the environment where the image acquisition device is located. Specifically, adjusting the resolution parameters of the image acquisition device can change the image size; adjusting the objective lens can change the image size range; adjusting the type of light source in the environment where the image acquisition device is located (e.g., switching from a bright field light source to a dark field light source) can change the image imaging effect; and adjusting the lighting conditions of the environment where the image acquisition device is located can change the image exposure effect.

[0164] In practical applications, the optical parameters can be adjusted multiple times, and the steps described in the above embodiments can be executed to obtain the relative pose of the image acquisition device focused on the acquisition object for each optical parameter. Then, according to the actual situation and needs, the optical parameters can be selected, and the corresponding relative pose of the image acquisition device focused on the acquisition object can be obtained for image acquisition.

[0165] It is understood that the embodiments described above provide multiple implementation schemes, and these implementation schemes can be combined and cross-referenced with each other without conflict, thereby extending to multiple possible implementation schemes. These can all be considered as the implementation schemes disclosed and made public in this specification.

[0166] This specification also provides an image acquisition control device corresponding to the above-described image acquisition control method. The following detailed description, with reference to the accompanying drawings, uses specific embodiments. It should be noted that the image acquisition control device described below can be considered as a functional module required to implement the image acquisition control method provided in this specification; the content of the image acquisition control device described below can be referred to in correspondence with the content of the image acquisition control method described above.

[0167] In specific implementation, such as Figure 4 The diagram shown is a structural block diagram of an image acquisition and control device provided in an embodiment of this specification. Figure 4 In this context, the image acquisition control device M10 can be connected to at least an image acquisition device (not shown in the figure), and specifically may include:

[0168] The image acquisition unit M11 is adapted to acquire images acquired by the image acquisition device at multiple relative poses when the relative pose between the image acquisition device and the acquisition object changes continuously.

[0169] The information acquisition unit M12 is adapted to determine imaging change information based on multiple acquired images;

[0170] The evaluation unit M13 is adapted to evaluate the imaging quality of the image acquisition device at each of the relative poses based on the imaging change information, obtain the imaging quality evaluation result, and determine the relative pose of the image acquisition device focusing on the acquisition object based on the imaging quality evaluation result.

[0171] By adopting the above scheme, since multiple images are acquired by the image acquisition device during the continuous change of relative pose, the influence of the surrounding environment on image acquisition can be reduced, and the timeliness of image acquisition can be improved. By evaluating the imaging quality of the image acquisition device through imaging change information, the obtained imaging quality evaluation results can accurately reflect the true focusing situation of the image acquisition device. Based on the imaging quality evaluation results, the changing trend of the true focusing situation of the image acquisition device can be obtained. Therefore, the relative pose that meets the expected focusing effect can be adaptively determined according to the actual situation. Thus, the image acquisition control method provided in this specification can ensure that the image acquisition device is in clear focus during image acquisition and can ensure the consistency of the focusing situation during image acquisition, effectively improving imaging quality and imaging stability, and thus improving the accuracy and reliability of image processing results.

[0172] Furthermore, the image acquisition and control device provided in the embodiments of this specification is applicable to application scenarios where there are at least one of the following: variable acquisition objects and batch image acquisition, thus having greater versatility.

[0173] In practice, the continuous change in the relative pose between the image acquisition device and the acquisition object can be achieved by the image acquisition control device or by other control devices.

[0174] When the image acquisition control device controls the continuous change of the relative pose between the image acquisition device and the acquisition object, the image acquisition control device may further include: a control unit adapted to control the continuous change of the relative pose between the image acquisition device and the acquisition object.

[0175] Furthermore, the specific implementation method of the relative pose change can be determined based on the actual connection relationship of the control unit. As an optional example, the control unit can be connected to the image acquisition device, so that when the continuous relative pose change is achieved, the control unit can control the image acquisition device to move continuously. As another optional example, the image acquisition control device can be connected to both the image acquisition device and the loading device for the acquisition object, thereby selectively controlling at least one of the image acquisition device and the acquisition object to move continuously, so that the relative pose changes continuously.

[0176] Furthermore, during the continuous change of the relative pose, the control device of the image acquisition control device can also control the image acquisition device to acquire multiple images of the acquisition object to obtain multiple images.

[0177] It is understandable that the speed and direction of continuous motion can be set according to specific application scenarios and requirements, as described in the relevant sections above, and will not be repeated here.

[0178] In specific implementations, the imaging change information may include at least one of the following: gradient change information; texture change information. The methods for obtaining the gradient change information and texture change information can be referred to the descriptions in the relevant sections above, and will not be repeated here.

[0179] In specific implementation, continue to refer to Figure 4 The evaluation unit M13 is adapted to perform a weighted operation on the gradient change information and texture change information corresponding to each relative pose after determining the gradient change information and texture change information, to obtain the corresponding imaging quality evaluation information, and to determine the imaging quality evaluation result based on each relative pose and the corresponding imaging quality evaluation information.

[0180] Therefore, by using weighted calculations, the importance of the gradient change information and the texture change information in the imaging quality assessment results can be flexibly controlled, thereby obtaining imaging quality assessment results with more reference value.

[0181] In specific implementation, continue to refer to Figure 4 The evaluation unit M13 is adapted to determine imaging change information multiple times based on multiple acquired images to obtain multiple imaging change information; based on each of the imaging change information, evaluate the imaging quality of the image acquisition device at each of the relative poses to obtain corresponding candidate quality evaluation results; based on each of the imaging change information, evaluate the imaging quality of the image acquisition device at each of the relative poses, match the multiple candidate quality evaluation results, and, after the matching result meets the confidence condition, determine the imaging quality evaluation result based on the multiple candidate quality evaluation results.

[0182] The matching process for multiple candidate quality assessment results and the judgment process for credibility conditions can be referred to the descriptions in the relevant sections above, and will not be repeated here.

[0183] In specific implementation, such as Figure 5 The diagram shown is a structural block diagram of another image acquisition and control device provided in an embodiment of this specification. Figure 5 In this embodiment, the image acquisition and control device M20 may include a memory M21 and a processor M22. The memory M21 stores computer instructions that can be executed on the processor M22. When the processor M22 executes the computer instructions, it can perform the steps of the image acquisition and control method provided in the embodiments of this specification.

[0184] The specific content and implementation of the image acquisition control method can be found in the descriptions of the relevant sections above, and will not be repeated here.

[0185] By adopting the above scheme, since multiple images are acquired by the image acquisition device during continuous changes in relative pose, the influence of the surrounding environment on image acquisition can be reduced, improving the timeliness of image acquisition. By evaluating the imaging quality of the image acquisition device through imaging change information, the obtained imaging quality evaluation results can accurately reflect the true focusing situation of the image acquisition device. Based on the imaging quality evaluation results, the changing trend of the true focusing situation of the image acquisition device can be obtained. Therefore, according to the actual situation, the relative pose that meets the expected focusing effect can be adaptively determined. Thus, the image acquisition control method provided in this specification can ensure that the image acquisition device has clear focus during image acquisition and can ensure the consistency of the focusing situation during image acquisition, effectively improving imaging quality and imaging stability, and thereby improving the accuracy and reliability of image processing results. Furthermore, the image acquisition control device provided in this specification is applicable to application scenarios where there are at least one of the following: variable acquisition objects and batch image acquisition, exhibiting higher versatility.

[0186] This specification also provides an image acquisition system corresponding to the image acquisition control method described above. The following detailed description, with reference to the accompanying drawings, uses specific embodiments to illustrate the system. It should be noted that the description of the image acquisition system below corresponds to the description of the image acquisition control device and image acquisition control method above.

[0187] In specific implementation, such as Figure 6 The diagram shown is a structural block diagram of an image acquisition system provided in an embodiment of this specification. Figure 6 In this context, the image acquisition system Z10 may include:

[0188] Loading device Z11 is suitable for loading the object to be collected;

[0189] Image acquisition device Z12, suitable for acquiring images of the acquisition object;

[0190] Image acquisition control device Z13 is connected to image acquisition device Z12; when the relative pose between the image acquisition device and the acquisition object changes continuously, it acquires images acquired by the image acquisition device at multiple relative poses; based on the acquired multiple images, it determines imaging change information; based on the imaging change information, it evaluates the imaging quality of the image acquisition device at each relative pose to obtain an imaging quality evaluation result; based on the imaging quality evaluation result, it determines the relative pose of the image acquisition device focusing on the acquisition object.

[0191] As can be seen from the above, acquiring multiple images during the continuous change of relative pose of the image acquisition device through the image acquisition control device can reduce the influence of the surrounding environment on image acquisition and improve the timeliness of image acquisition. By evaluating the imaging quality of the image acquisition device through imaging change information, the obtained imaging quality evaluation results can accurately reflect the true focusing situation of the image acquisition device. Based on the imaging quality evaluation results, the changing trend of the true focusing situation of the image acquisition device can be obtained. Therefore, according to the actual situation, the relative pose that meets the expected focusing effect can be adaptively determined. Thus, the image acquisition control method provided in this specification can ensure that the image acquisition device has clear focus during image acquisition and can ensure the consistency of the focusing situation during image acquisition, effectively improving imaging quality and imaging stability, and thereby improving the accuracy and reliability of image processing results.

[0192] In practical implementation, the relative pose change between the image acquisition device and the loading device can be achieved through a pose adjustment device. It is understood that the specific components of the pose adjustment device can be determined according to the actual application scenario and requirements, thereby forming a mechanical structure capable of achieving at least one adjustment function: position adjustment and angle adjustment.

[0193] For example, the posture adjustment device may include a piston assembly and a piston controller. The image acquisition device is connected to the movable axis of the piston assembly. The piston controller controls the movement speed and direction of the piston assembly, enabling the image acquisition device to perform linear movement.

[0194] For example, the posture adjustment device may include a motor and a motor controller. The image acquisition device is connected to the rotating shaft of the motor. The motor controller controls the speed and direction of rotation of the motor, which can rotate the angle of the image acquisition device.

[0195] In practical implementation, the specific image processing steps performed by the image processing device can be set according to the application scenario and requirements. For example, in detection scenarios such as defect detection, critical dimension detection, and key point detection, the image processing device can be set to perform image processing steps related to defect detection, critical dimension detection, and key point detection; in target recognition scenarios, the image processing device can be set to perform image processing steps related to target recognition. This specification does not impose specific limitations on the steps performed by the image processing device.

[0196] This specification also provides a computer-readable storage medium storing computer instructions that, when executed, perform the steps of the image acquisition control method described in any of the foregoing embodiments. Specific steps can be referred to the methods described in the foregoing embodiments, and will not be repeated here.

[0197] The computer-readable storage medium may include memory, removable or non-removable medium, erasable or non-erasable medium, writable or rewritable medium, digital or analog medium, etc.

[0198] Computer instructions may include any suitable type of code implemented using any appropriate high-level, low-level, object-oriented, visual, compiled, and / or interpreted programming language, such as source code, compiled code, interpreted code, executable code, static code, dynamic code, encrypted code, etc.

[0199] While the embodiments disclosed in this specification are as described above, they are not limited thereto. Any person skilled in the art can make various modifications and alterations without departing from the spirit and scope of the embodiments in this specification. Therefore, the scope of protection of the embodiments in this specification should be determined by the scope defined in the claims.

Claims

1. An image acquisition control method, characterized in that, include: When the relative pose between the image acquisition device and the acquisition object changes continuously, the image acquisition device acquires images at multiple relative poses respectively. Based on multiple acquired images, imaging change information is determined, including: determining a reference image and a corresponding historical image from the multiple images, and determining the texture change information of each reference image relative to the historical image; wherein, determining the texture change information includes: taking each reference image and each historical image as images to be processed, selecting multiple pixel groups around a specified pixel point in the image to be processed according to a specified direction, and determining the gradient of each pixel group of the specified pixel point, and determining the contour features of the image to be processed based on the gradient of each pixel group of the specified pixel point; determining the contour region in the image to be processed based on the extracted contour features, and extracting the texture information of the image to be processed by region gradient feature extraction based on the contour region in the image to be processed, and determining the texture change information by comparing the texture information obtained by taking the reference image as the image to be processed and the texture information obtained by taking the historical image as the image to be processed; Based on the imaging change information, the imaging quality of the image acquisition device at each of the relative poses is evaluated to obtain the imaging quality evaluation result. Based on the imaging quality assessment results, the relative pose of the image acquisition device focusing on the acquisition object is determined.

2. The image acquisition control method according to claim 1, characterized in that, The continuous change in relative pose between the image acquisition device and the acquisition object includes: Control at least one of the image acquisition device and the acquisition object to perform continuous movement so that the relative pose changes continuously.

3. The image acquisition control method according to claim 1, characterized in that, The determination of imaging change information based on multiple acquired images also includes: Determine the gradient change information for each of the images.

4. The image acquisition control method according to claim 3, characterized in that, Determining the gradient change information of each of the images includes: Extract the gradient features of each image and determine the gradient change information.

5. The image acquisition control method according to claim 4, characterized in that, The extraction of gradient features from each of the images includes: Gradient features of each image are extracted based on the multi-point gradient operator of adjacent matrices.

6. The image acquisition control method according to claim 3, characterized in that, Determining the gradient change information of each of the images includes: The gradient change information is determined based on a specified region in each of the images.

7. The image acquisition control method according to claim 1, characterized in that, The step of selecting multiple pixel groups around a specified pixel in the image to be processed according to a specified direction, determining the gradient of each pixel group of the specified pixel, and determining the contour features of the image to be processed based on the gradients of each pixel group of the specified pixel, includes: According to the first direction, multiple first pixel groups are selected around a specified pixel in the image to be processed, and after calculating the gradient of each first pixel group corresponding to the specified pixel, the contour features of the image to be processed in the first direction are determined based on the gradient of each first pixel group of the specified pixel. In accordance with the second direction, multiple second pixel groups are selected around a specified pixel in the image to be processed, and after calculating the gradient of each second pixel group corresponding to the specified pixel, the contour features of the image to be processed in the second direction are determined based on the gradient of each second pixel group of the specified pixel. Based on the contour features of the image to be processed in the first direction and the contour features in the second direction, the contour features of the image to be processed are determined.

8. The image acquisition control method according to claim 1 or 7, characterized in that, Before performing a weighted calculation on the gradient of each pixel group of the specified pixel point, the method further includes: Based on the gradient change rate among the multiple pixel groups of the specified pixel point, a contour weight is assigned to the gradient of each pixel group.

9. The image acquisition control method according to claim 3, characterized in that, Determining the texture change information of each of the reference images relative to the historical images includes: Texture change information is determined based on specified regions in each of the reference images and specified regions in each of the historical images.

10. The image acquisition control method according to claim 3, characterized in that, The process of evaluating the imaging quality of the image acquisition device at each relative pose based on the imaging change information, and obtaining imaging quality evaluation results, includes: The gradient change information and texture change information corresponding to each relative pose are weighted and calculated to obtain the corresponding imaging quality evaluation information. Based on the relative poses and corresponding imaging quality assessment information, the imaging quality assessment result is determined.

11. The image acquisition control method according to claim 1, characterized in that, The determination of imaging change information based on multiple acquired images includes: Based on multiple acquired images, imaging change information is determined multiple times to obtain multiple imaging change information. The process of evaluating the imaging quality of the image acquisition device at each relative pose based on the imaging change information, and obtaining imaging quality evaluation results, includes: Based on the imaging change information, the imaging quality of the image acquisition device at each of the relative poses is evaluated to obtain the corresponding candidate quality evaluation results. Match multiple candidate quality assessment results; After the matching result meets the confidence criteria, the imaging quality assessment result is determined based on the multiple candidate quality assessment results.

12. An image acquisition and control device, characterized in that, Connects to an image acquisition device, including: An image acquisition unit is adapted to acquire images acquired by the image acquisition device at multiple relative poses when the relative pose between the image acquisition device and the acquisition object changes continuously. An information acquisition unit is adapted to determine imaging change information based on multiple acquired images, including: determining a reference image and a corresponding historical image from the multiple images respectively, and determining the texture change information of each reference image relative to the historical image; wherein, determining the texture change information includes: taking each reference image and each historical image as images to be processed, selecting multiple pixel groups around a specified pixel point in the images to be processed according to a specified direction, and determining the gradient of each pixel group of the specified pixel point, and determining the contour features of the images to be processed based on the gradient of each pixel group of the specified pixel point; determining the contour region in the images to be processed based on the extracted contour features, and extracting the texture information of the images to be processed by region gradient feature extraction based on the contour region in the images to be processed, and determining the texture change information by comparing the texture information obtained by taking the reference image as the images to be processed with the texture information obtained by taking the historical image as the images to be processed; The evaluation unit is adapted to evaluate the imaging quality of the image acquisition device at each of the relative poses based on the imaging change information, obtain the imaging quality evaluation result, and determine the relative pose of the image acquisition device focusing on the acquisition object based on the imaging quality evaluation result.

13. An image acquisition and control device, comprising a memory and a processor, wherein the memory stores computer instructions executable on the processor, characterized in that, When the processor executes the computer instructions, it performs the steps of the method according to any one of claims 1 to 11.

14. An image acquisition system, characterized in that, include: Loading equipment, suitable for loading the objects to be collected; An image acquisition device, suitable for acquiring images of the object to be acquired; Image acquisition control equipment, connected to the image acquisition equipment; It is suitable for acquiring images acquired by the image acquisition device at multiple relative poses when the relative pose between the image acquisition device and the acquisition object changes continuously; Based on multiple acquired images, determining imaging change information includes: determining a reference image and a corresponding historical image from the multiple images, and determining the texture change information of each reference image relative to the historical image; evaluating the imaging quality of the image acquisition device at each relative pose based on the imaging change information to obtain an imaging quality evaluation result; determining the relative pose of the image acquisition device focused on the acquired object based on the imaging quality evaluation result, wherein determining the texture change information includes: taking each reference image and each historical image as images to be processed, selecting multiple pixel groups around a specified pixel point in the image to be processed according to a specified direction, and determining the gradient of each pixel group of the specified pixel point, and determining the contour features of the image to be processed based on the gradient of each pixel group of the specified pixel point; determining the contour region in the image to be processed based on the extracted contour features, and extracting the texture information of the image to be processed by region gradient feature extraction based on the contour region in the image to be processed, and determining the texture change information by comparing the texture information obtained by taking the reference image as the image to be processed with the texture information obtained by taking the historical image as the image to be processed.

15. A computer-readable storage medium storing computer instructions thereon, characterized in that, When the computer instructions are executed, they perform the steps of the method according to any one of claims 1 to 11.

Images