A method and system for visual positioning of a bottle body

Abstract

This invention relates to the field of visual inspection, and more particularly to a method and system for visual positioning of a bottle. The method is used to perform pose positioning of a transparent bottle. First, in response to the bottle's positioning, a pre-configured ring light source is activated to provide supplementary illumination and a photographing command. Then, in response to the triggering of the ring light source and photographing command, an image of the bottle is acquired. Next, the bottle image undergoes de-reflection and enhancement processing to obtain a feature-enhanced image. Then, feature extraction is performed on the feature-enhanced image to obtain multi-scale features containing the bottle's outline and local features. These multi-scale features are matched with preset template features. If the matching result is satisfactory, abnormal features in the multi-scale features are removed to obtain corrected features. Finally, the corrected features are input into a preset bottle pose algorithm to obtain pose parameters that are applied to downstream devices for alignment. Compared to existing technologies, this invention achieves precise positioning of transparent bottles.

Description

Technical Field

[0001] This invention relates to the field of visual inspection technology. More specifically, this invention relates to a method and system for visual positioning of a bottle. Background Technology

[0002] Bottle visual positioning refers to the technique of using visual inspection technology to locate the coordinates of key feature points on the bottle before screen printing. The accuracy of this positioning directly affects the subsequent screen printing quality. To achieve the positioning of key feature points on the bottle, Chinese patent application CN106875441A discloses an online PE bottle recognition and positioning method based on machine vision. It first uses two cameras to acquire images of the PE bottle opening and body, and performs grayscale transformation on the initial images. Then, it preprocesses the bottle opening and body images using median filtering and Laplacian sharpening. Next, it uses the Canny operator to extract the bottle body contour and bottle opening contour. Then, it calculates the pixel points of the bottle body contour to complete PE bottle target recognition based on contour perimeter matching. Finally, it uses the least squares method to fit an ellipse to the discrete points of the image contour, and calculates the centroid coordinates of the ellipse based on the fitted ellipse equation, using these coordinates as the pixel coordinates of the PE bottle opening.

[0003] However, the above technical solution still has the following technical defects, specifically: First, the above solution is only applicable to the detection of PE bottles. PE bottles are mainly made of opaque and matte materials. When acquiring visual images, the images will not reflect light. If the above solution is directly used to detect transparent and textured bottles, it will be difficult to identify key feature points in the image.

[0004] Secondly, the slight rotation or displacement of the robotic arm when grasping the bottle will exacerbate the difficulty in recognizing bottle features and cause positioning deviations, ultimately failing to meet the high-precision requirements of printing alignment.

[0005] Therefore, the main problem with existing technologies is the difficulty in accurately positioning transparent material bottles. Summary of the Invention

[0006] To address the aforementioned technical problem of accurately positioning transparent material bottles, this invention discloses a method and system for visual positioning of bottles.

[0007] In a first aspect, the present invention discloses a method for visual positioning of a bottle body, used for posing positioning of a transparent bottle body. The method of the present invention includes: In response to the bottle's positioning, the system activates a pre-configured ring light source for illumination and triggers a photographing command. In response to the ring light source supplementary lighting and the triggering of the photo-taking command, an image of the bottle is captured; The bottle image is de-reflected and enhanced to obtain a feature-enhanced image; Feature extraction is performed on the feature-enhanced image to obtain multi-scale features that include the bottle outline and local features; Match multi-scale features with preset template features; If the matching result is satisfactory, abnormal features in the multi-scale features are removed to obtain the corrected features; The corrected features are input into the preset bottle pose algorithm to obtain pose parameters, which are then applied to downstream devices to perform alignment.

[0008] Beneficial Effects: In bottle positioning, the method of this invention first directly drives a ring light source for supplementary lighting, utilizing the principle of ring diffuse reflection to eliminate some of the reflective interference from the transparent bottle, thereby improving the contrast and recognizability of the bottle image. Then, after the aforementioned hardware-level pre-processing for reflection, the bottle image undergoes de-reflection and enhancement post-processing to obtain a feature-enhanced image. Next, feature extraction is performed on the feature-enhanced image to obtain multi-scale features to support the image matching process in subsequent steps. After successful image matching, abnormal features are removed to eliminate positioning interference factors. Finally, the corrected features are input into the bottle pose algorithm to obtain pose parameters, thereby achieving precise positioning of the transparent bottle. Compared to existing technologies, this invention achieves precise positioning of the transparent bottle by introducing hardware and software reflection elimination strategies, multi-scale feature matching strategies, and abnormal feature removal strategies.

[0009] Preferably, the ring light source supplementary lighting and photo capture command are configured as follows: Read the camera parameters and lighting parameters; the camera parameters include visual rotation speed, visual positioning circle count, visual jog speed, and lens exposure time; the lighting parameters include light source illumination time, light intensity, and illumination angle. The camera takes pictures using pre-configured camera parameters; The pre-configured ring light source is driven by the fill light parameters to provide fill light.

[0010] Preferably, the ring light source supplementary lighting and photo capture command are also configured as follows: Read the polarization angle; The pre-configured polarization unit is driven by the polarization angle.

[0011] Preferably, the image of the bottle is subjected to de-reflection and enhancement processing, including: Read the material properties corresponding to the bottle image; Based on the material properties, retrieve the corresponding specular threshold from the database; Based on the highlight threshold, the highlight mask of the bottle image is extracted and segmented to obtain the highlight image and the image to be repaired. The fast traversal algorithm is used to repair the image to be repaired, resulting in the first repaired image; Histogram equalization is performed on the first restored image to obtain a feature-enhanced image.

[0012] Preferably, before matching the multi-scale features with the preset template features, this method further includes: Configure multiple sets of bottle outline edge templates of different categories and store them in the corresponding template database; Configure the pyramid hierarchical search script to serve as a search constraint for the template database.

[0013] Preferably, the pyramid hierarchical search script is configured as follows: Read the unique search identifier corresponding to the bottle image; Locate the template database for the corresponding group based on the unique search identifier; In the template database, bottle outline edge templates are retrieved according to the rules of extraction from low resolution to high resolution to support feature matching.

[0014] Preferably, matching multi-scale features with preset template features includes: Retrieve the corresponding bottle outline template; Feature extraction is performed on the bottle outline edge template to obtain template features; The similarity is calculated by matching the multi-scale features with the template features. Determine if the similarity is higher than the threshold; if so, it is considered acceptable.

[0015] Preferably, if the matching result is satisfactory, this method further includes: The optimal matching position of the bottle image is fitted with a quadratic surface, and the multi-scale features are updated.

[0016] Preferably, the bottle pose algorithm adopts the PnP pose algorithm.

[0017] In a second aspect, the present invention also discloses a system for visual positioning of a bottle, including a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the method for visual positioning of a bottle described in the first aspect is implemented.

[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The method of the present invention achieves the effect of precise positioning of transparent bottle body by introducing hardware and software reflection elimination strategy, multi-scale feature matching strategy and abnormal feature removal strategy.

[0019] (2) The method of the present invention introduces a lens, a ring light source and a polarization unit to eliminate image reflection at the physical level, which has high reliability.

[0020] (3) The method of the present invention introduces a pyramid hierarchical search strategy to achieve efficient matching. Attached Figure Description

[0021] The above and other objects, features, and advantages of exemplary embodiments of the present invention will become readily apparent upon reading the following detailed description with reference to the accompanying drawings. In the drawings, several embodiments of the invention are illustrated by way of example and not limitation, and like or corresponding reference numerals denote like or corresponding parts, wherein: Figure 1 This is a flowchart of the method for visual positioning of the bottle body in Embodiment 1 of the present invention; Figure 2 This is an output effect diagram of the pose parameters in Embodiment 1 of the present invention; Figure 3 This is a schematic diagram of the system structure for visual positioning of the bottle body in Embodiment 2 of the present invention. Detailed Implementation

[0022] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0023] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0024] Example 1 like Figure 1 As shown, this invention discloses a method for visual positioning of a bottle, comprising: S10: In response to the bottle's positioning, it activates a pre-configured ring light source for illumination and takes a picture.

[0025] In this embodiment, the method is mainly used for pose localization of a transparent bottle. In other embodiments, if it is necessary to detect other types of bottles, only the image acquisition and template matching functions need to be retained.

[0026] Furthermore, the aforementioned ring light source supplementary lighting and photo-taking instructions are configured as follows: S100: Reads camera parameters and fill light parameters.

[0027] The camera parameters include visual rotation speed, visual positioning rotation number, visual jog speed, and lens exposure time; the lighting parameters include light source illumination time, light intensity, and illumination angle.

[0028] It should be explained that the aforementioned visual rotation speed refers to the speed at which the visual system perceives a rotating object, preferably 3.5 r / s-4.5 r / s. The visual positioning cycle count refers to the number of cycles the visual system uses for precise positioning during calibration or recognition, preferably 1.5-2.5 cycles. The visual jog speed refers to the speed at which the displacement actuator of the visual camera performs small, intermittent movements controlled by visual feedback, used to achieve high-precision position adjustment and positioning; it is a dynamically changing value and needs to be configured according to the expected image resolution. The lens exposure time refers to the length of time the visual camera shutter is open, used to determine the duration for which the camera's photosensitive element receives light; it should be noted that it cannot exceed 0.5 seconds. The start time of the light source ignition time must be earlier than the start time of the lens exposure time, and the end time of the light source ignition time must be later than the end time of the lens exposure time. The light intensity is best controlled between 800 lux and 2000 lux. With the vertical line from the plane of the bottle directly illuminated by the light source as the 0° reference, the illumination angle is preferably 0°-35°.

[0029] S101: Drives the pre-configured camera to take pictures using camera parameters.

[0030] S102: Drive a pre-configured ring light source to provide supplementary lighting using supplementary lighting parameters.

[0031] It should be noted that steps S101-S102 are executed simultaneously. After reading the camera parameters and the supplementary lighting parameters, the industrial control computer drives the corresponding actuators to take pictures and perform supplementary lighting. Specifically, the actuator in step S101 mainly consists of a vision camera and a vision camera displacement mechanism; the actuator in step S102 mainly consists of a ring light source and its angle adjustment mechanism.

[0032] Through the above steps S101-S102, the method of this embodiment can directly drive the ring light source for supplementary lighting and simultaneously take pictures with the vision camera, thereby using the ring diffuse reflection principle to eliminate part of the reflection interference of the transparent bottle body, thereby improving the contrast and recognizability of the bottle body image.

[0033] Furthermore, to further reduce the decrease in image recognition accuracy caused by reflection, this embodiment also introduces a polarization control strategy. Specifically, a polarizer can be installed in front of the ring light source to emit polarized light in a specific direction to illuminate the surface of the transparent bottle; simultaneously, a rotatable polarizing mirror is installed in front of the lens of the vision camera to selectively receive or block reflected polarized light. By adjusting the polarization angle of the two polarization units, specular reflection can be effectively suppressed and surface detail contrast enhanced, thereby obtaining a clearer image. Therefore, the aforementioned ring light source supplementary lighting and image capture command can also be configured to first read the polarization angle and then drive the pre-configured polarization unit with the polarization angle, thereby realizing its application in industrial inspection.

[0034] S20: In response to the ring light source supplementary lighting and the triggering of the photo-taking command, it captures an image of the bottle.

[0035] In one embodiment, to improve positioning efficiency, only one bottle image can be acquired for a single sample.

[0036] In another embodiment, to improve positioning accuracy, multiple consecutive frames of bottle images can be acquired for a single sample.

[0037] S30: Perform anti-reflection and enhancement processing on the bottle image to obtain a feature-enhanced image.

[0038] Specifically, step S30 above includes: S31: Read the material properties corresponding to the bottle image.

[0039] It should be noted that the above material properties can be glass, PET plastic, PP plastic, PVC plastic, and polymer resin.

[0040] S32: Based on the material properties, retrieve the corresponding specular threshold from the database.

[0041] It's important to clarify that the specular threshold is a dynamic, adaptive threshold. Different materials have different reflective properties, requiring the specular threshold to be configured according to the specific situation. Therefore, the adaptive change of the specular threshold can be achieved by pre-configuring the mapping relationship between materials and specular thresholds.

[0042] S33: Extract the highlight mask from the bottle image based on the highlight threshold, and segment it to obtain the highlight image and the image to be repaired.

[0043] Specifically, the acquired color or grayscale image of the bottle is converted into a luminance channel image. Since the pixel grayscale values ​​of the specular reflection area are usually concentrated in the extremely high frequency range, the aforementioned specular threshold can be used to locate and segment the reflective area.

[0044] S34: The fast traversal algorithm is used to repair the image to be repaired, and the first repaired image is obtained.

[0045] Specifically, after removing the reflective areas, the original texture information of the highlight areas has been optically lost, making it impossible to directly restore the features. Therefore, this embodiment introduces the Fast Marching Model (FMM) algorithm to fill in the missing reflective areas.

[0046] More specifically, using the edge pixels of the highlight threshold as the boundary, a fast traversal algorithm is used to progressively advance from the boundary into the reflective area. During this advancement, the grayscale gradient and normal direction of known pixels (non-reflective pixels) near the boundary are calculated. A weighted average method is then used to interpolate and fill the surrounding normal bottle texture and color towards the reflective center. After restoration, the originally glaring reflective spots are smoothly replaced with a transitional background consistent with the surrounding bottle, thus completing the algorithmic "de-reflection" process and obtaining the initial restored image.

[0047] S35: Perform histogram equalization on the first repaired image to obtain a feature-enhanced image.

[0048] Specifically, after the initial restoration is completed, due to uneven lighting on the curved surface of the bottle, the edge areas are often darker, and the overall image contrast may decrease. To highlight the details to be detected (such as anti-counterfeiting codes and minor scratches), Limit Contrast Adaptive Histogram Equalization (CLAHE) is performed on the initial restored image. More specifically, in the specific execution of step S35, the first restored image needs to be divided into N... N non-overlapping local image patches are identified. Then, the gray-level histogram of each image patch is calculated, and a contrast limit threshold is set. If the number of pixels at certain gray levels in any image patch exceeds the limit, the excess pixels are uniformly cropped and redistributed to other gray levels to prevent background noise from being amplified due to over-enhancement. Finally, after histogram equalization of all image patches, the boundaries of adjacent image patches are smoothly stitched together using a bilinear interpolation algorithm to eliminate block artifacts and obtain the final feature-enhanced image.

[0049] Through the above technical steps S31-S35, this embodiment of the method also introduces a post-processing procedure for reflection of the image processing layer, which improves the anti-reflection interference capability of this embodiment of the method and is particularly suitable for image recognition of transparent bottles.

[0050] S40: Perform feature extraction on the feature-enhanced image to obtain multi-scale features that include the bottle outline and local features.

[0051] It needs to be explained that, for example Figure 2 As shown, the bottle outline mainly refers to the bottle edge, while local features include identifying features such as labels, dents / embossings, or barcodes on the bottle. The Canny operator can be used to extract these bottle edges and their identifying features. The extracted bottle edges and their identifying features must also meet area and perimeter constraints. Furthermore, after extracting the bottle edges and their identifying features, contour fitting is required to form a closed and smooth curve.

[0052] S50: Match multi-scale features with preset template features.

[0053] It should be noted that before performing step S50 above, the following operations need to be performed on the template database: S500: Configure multiple sets of bottle outline edge templates of different categories and store them in the corresponding template database.

[0054] S501: Configure the pyramid hierarchical search script as a search constraint for the template database.

[0055] The pyramid hierarchical search script is configured to first read the unique search identifier corresponding to the bottle image, and then locate the template database of the corresponding group based on the unique search identifier. Then, when it is necessary to retrieve an image from the template database for matching, the bottle outline edge template used to support feature matching is retrieved according to the rule of extraction from low resolution to high resolution.

[0056] The efficiency of the method in this embodiment is effectively improved through the above-mentioned template search configuration optimization.

[0057] Specifically, after completing the above-mentioned search configuration for the template database, when executing step S50, the method of this embodiment includes: S51: Retrieve the corresponding bottle outline edge template.

[0058] In this embodiment, the batch number or material attribute can be generated by first reading the bottle image, and then the bottle outline edge template can be retrieved from the template database of the classification number.

[0059] S52: Extract features from the bottle outline edge template to obtain template features.

[0060] Regarding feature extraction, the objects of feature extraction need to be consistent with multi-scale features, that is, including contours and corresponding local feature extraction.

[0061] It should be noted that, in a preferred strategy, a visual interaction system can be used to first select template features containing the outline and corresponding local features on the bottle outline edge template, and then store them together in the template database. This can reduce the recognition burden of the visual system and improve the overall efficiency.

[0062] S53: Perform similarity matching between multi-scale features and template features to calculate the similarity.

[0063] In this embodiment, the algorithm expression for the above similarity matching is:

[0064] In the formula, This represents the similarity score, and its value ranges from 0 to 1. Feature dimensions representing multi-scale features; The first image on the bottle One feature parameter; The first part represents the template of the bottle outline edge. One feature parameter; from Returns the maximum value among the feature matching results; cos represents the cosine function.

[0065] S54: Determine if the similarity is higher than the threshold. If so, it is considered qualified.

[0066] Through steps S51-S53 described above, coarse localization based on template matching can be achieved. Next, the vision system selects the best matching position that meets the requirements, performs quadratic surface fitting on the best matching position of the bottle image, and updates the multi-scale features.

[0067] S60: If the matching result is qualified, remove the abnormal features in the multi-scale features to obtain the corrected features.

[0068] In this embodiment, the method introduces the RANSAC (Random Sample Consensus) algorithm to eliminate abnormal matching features, thereby improving the robustness of feature matching. In this way, all factors affecting accurate localization are comprehensively eliminated.

[0069] S70: Input the corrected features into the preset bottle pose algorithm to obtain pose parameters, which are then applied to the downstream device to perform alignment.

[0070] It should be noted that each bottle outline edge template also needs to be pre-configured with 3D-2D point pairs, i.e., the mapping relationship between pixels in the two-dimensional image and points in the three-dimensional space. After completing the matching and coarse localization, the corrected features themselves possess the basic information (coordinates) of the pixels in the two-dimensional image. Using the matching mapping relationship, the corresponding 3D point pairs are obtained. Therefore, when the corrected features are input into the preset bottle pose algorithm, the 3D point pairs matched by the corrected features will be automatically obtained and input into the bottle pose algorithm as well.

[0071] In this embodiment, to achieve accurate positioning, the bottle pose algorithm described above employs the PnP pose algorithm. It should be noted that the PnP pose algorithm is a common algorithm used in computer vision to solve for the 3D pose of a camera or object.

[0072] Intuitively, the PnP pose algorithm uses already matched 3D-2D point pairs to solve the problem, ultimately obtaining the following... Figure 2 The bottle body pose parameters shown mainly include the coordinates of key features (identifiers) and the tilt angle of the bottle body relative to the standard template.

[0073] Downstream equipment can be screen printing, transfer printing, hot stamping, or other devices that require reading printing alignment data (pose parameters). Through the above technical solution, the pose parameters output by the method of this invention enable downstream equipment to output products of higher quality.

[0074] Compared with the prior art, the method of this embodiment achieves performance improvements in at least the following dimensions through the above technical solution: Firstly, regarding the pose parameters of the transparent bottle, the positioning accuracy in coordinates has been improved from 1mm to 0.07mm.

[0075] Secondly, the positioning time for each transparent bottle has been reduced from 170ms to 100ms.

[0076] Third, the success rate of identifying transparent bottles has increased from 85% to 99.5%.

[0077] Example 2 like Figure 3 As shown, this embodiment also discloses a system for body visual positioning, including a processor and a memory. The memory stores computer program instructions, which, when executed by the processor, implement the method for bottle body visual positioning described in the first aspect.

[0078] The system in this embodiment also includes other components well known to those skilled in the art, such as communication interfaces. Their settings and functions are known in the art, and therefore will not be described in detail here.

[0079] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions that can be stored or otherwise maintained by such a computer-readable medium.

[0080] In the description of this specification, "multiple" means at least two, such as two, three or more, etc., unless otherwise expressly and specifically defined.

[0081] While this specification has shown and described numerous embodiments of the invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Many modifications, alterations, and alternatives will occur to those skilled in the art without departing from the spirit and essence of the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in the practice of this invention.

Claims

1. A method for visual positioning of a bottle, characterized in that, The method for posing and positioning a transparent bottle includes: In response to the bottle's positioning, the system activates a pre-configured ring light source for illumination and triggers a photographing command. In response to the ring light source supplementary lighting and the triggering of the photo-taking command, an image of the bottle is captured; The bottle image is subjected to de-reflection and enhancement processing to obtain a feature-enhanced image; Feature extraction is performed on the enhanced image to obtain multi-scale features that include the bottle outline and local features; The multi-scale features are matched with preset template features; Abnormal features that fail to match in the multi-scale features are removed to obtain corrected features; The modified features are input into a preset bottle pose algorithm to obtain pose parameters, which are then applied to downstream devices to perform alignment.

2. The method for visual positioning of a bottle body according to claim 1, characterized in that, The ring light source supplementary lighting and photo capture command are configured as follows: Read camera parameters and lighting parameters; wherein, the camera parameters include visual rotation speed, visual positioning circle count, visual jog speed, and lens exposure time; the lighting parameters include light source illumination time, light intensity, and illumination angle; The camera is driven to take pictures using the aforementioned camera parameters; The pre-configured ring light source is driven to provide supplementary lighting using the aforementioned supplementary lighting parameters.

3. The method for visual positioning of a bottle body according to claim 1, characterized in that, The ring light source supplementary lighting and photo capture command are also configured as follows: Read the polarization angle; The pre-configured polarization unit is driven at the polarization angle.

4. The method for visual positioning of a bottle body according to claim 1, characterized in that, The bottle image undergoes de-reflection and enhancement processing, including: Read the material properties corresponding to the bottle image; Based on the material properties, retrieve the corresponding specular threshold from the database; Based on the specified highlight threshold, the bottle image is extracted using a highlight mask, and segmented to obtain a highlight image and an image to be repaired. The image to be repaired is repaired using a fast traversal algorithm to obtain a first repaired image; Histogram equalization is performed on the first repaired image to obtain the feature-enhanced image.

5. The method for visual positioning of a bottle according to claim 1, characterized in that, Before matching the multi-scale features with the preset template features, the method further includes: Configure multiple sets of bottle outline edge templates of different categories and store them in the corresponding template database; Configure the pyramid hierarchical search script as a search constraint for the template database.

6. The method for visual positioning of a bottle body according to claim 5, characterized in that, The pyramid hierarchical search script is configured as follows: Read the unique search identifier corresponding to the bottle image; The template database corresponding to the group is located based on the unique search identifier; In the template database, bottle outline edge templates are retrieved according to the rules of extraction from low resolution to high resolution to support feature matching.

7. The method for visual positioning of a bottle body according to claim 5, characterized in that, Matching the multi-scale features with preset template features includes: Retrieve the corresponding bottle outline template; Feature extraction is performed on the bottle outline edge template to obtain template features; The similarity is calculated by matching the multi-scale features with the template features. Determine whether the similarity is higher than the threshold; if so, it is considered qualified.

8. The method for visual positioning of a bottle according to claim 1, characterized in that, If the matching result is satisfactory, the method further includes: The optimal matching position of the bottle image is fitted with a quadratic surface, and the multi-scale features are updated.

9. The method for visual positioning of a bottle according to claim 1, characterized in that, The bottle pose algorithm uses the PnP pose algorithm.

10. A system for visual positioning of a bottle, characterized in that, It includes a processor and a memory, the memory storing computer program instructions that, when executed by the processor, implement the method for visual positioning of a bottle as described in any one of claims 1-9.

Images