Document identification system

Abstract

The utility model provides a document identification system, which comprises an image acquisition device and a control device, the control device is configured to identify document information based on an image, and the image acquisition device comprises a box body; the paper pressing mechanism is fixed in the box body and is used for flattening and keeping the documents at the paper pressing mechanism; the camera is fixed towards the paper pressing mechanism in the box body so as to acquire an image of the document at the paper pressing mechanism, and the image is transmitted to the control device; the light source mechanism comprises a light-emitting device and a light shield, the light-emitting device is fixed in the light shield, and the light shield covers the paper pressing mechanism from the outside and is used for providing a lighting environment for shooting. By means of the document identification system and method, the workload of workers can be remarkably reduced, and high identification accuracy is achieved.

Description

Technical Field

[0001] The utility model relates to the technical field of automobiles, in particular to a document identification system and a document identification method which can be executed accordingly. Background Art

[0002] At present, in the process of vehicle manufacturing, the inspection of documents is mostly done by visual verification. For example, for each vehicle, the staff usually needs to check the vehicle identification number (VIN code, which contains 17 digits and letters) six times and the engine number (which contains 9 digits) four times during the entire manufacturing process. This poses a great challenge to the staff's operational ability, and it is easy to miss or mischeck when the staff is tired. In addition, each vehicle is also equipped with a variety of other identity information, so intelligent verification and accurate traceability of documents are extremely necessary.

[0003] It should be noted that the content introduced here only provides background information related to the present disclosure and does not necessarily belong to the prior art. Utility Model Content

[0004] According to different aspects, the purpose of the present utility model is to provide a document identification system and a document identification method, which can be used in the whole vehicle manufacturing process.

[0005] In addition, the utility model is also intended to solve or alleviate other technical problems existing in the prior art.

[0006] The utility model solves the above problems by providing a document recognition system. Specifically, the system comprises an image acquisition device and a control device. The control device is configured to recognize document information based on an image, and the image acquisition device comprises:

[0007] Box;

[0008] A paper pressing mechanism, which is fixed in the box and is used to flatten and hold the document therein;

[0009] a camera, which is fixed in the box toward the paper pressing mechanism to obtain an image of the document at the paper pressing mechanism, and the image is transmitted to the control device;

[0010] The light source mechanism comprises a light emitting device and a light shield, wherein the light emitting device is fixed inside the light shield, and the light shield covers a paper pressing mechanism on the outside and is used to provide a lighting environment for shooting.

[0011] In the document identification system proposed according to the first aspect of the utility model, the light shield is dome-shaped, and a light-emitting device is fixed on the inner wall of the light shield by means of a slide rail mechanism, and the light-emitting device is pushed into or pulled out of the light shield by the slide rail mechanism.

[0012] In the document identification system proposed according to the first aspect of the utility model, the paper pressing mechanism includes an upper light-transmitting plate and a lower light-transmitting plate placed flat against each other, wherein the upper light-transmitting plate can rotate around its edge fixed to the lower light-transmitting plate by means of a rotating bearing, so that the paper pressing mechanism can switch between an open state and a pressed state, in which the document can be placed between the upper light-transmitting plate and the lower light-transmitting plate, and in which the document is pressed tightly between the upper light-transmitting plate and the lower light-transmitting plate.

[0013] In the document identification system proposed according to the first aspect of the utility model, the upper light-transmitting plate or the lower light-transmitting plate is light-transmitting frosted glass.

[0014] In the document identification system proposed according to the first aspect of the utility model, the paper pressing mechanism also includes a handle mechanism, which is fixed to the upper light-transmitting plate and extends out of the box body with an operating handle, wherein a snap-fit ​​portion is provided on the operating handle, and the snap-fit ​​portion can be fixed to the outer wall of the box body in the opened state.

[0015] In the document recognition system proposed according to the first aspect of the utility model, cameras are provided above the upper light-transmitting plate and below the lower light-transmitting plate to capture images of the front and back of the document.

[0016] In the document identification system according to the first aspect of the utility model, a heat dissipation slot is provided on the box body.

[0017] In the document identification system proposed according to the first aspect of the utility model, a fan is fixed on the box to perform convection heat dissipation inside the box.

[0018] In the document identification system proposed according to the first aspect of the utility model, a filter is also provided at the fan to filter the air entering the box.

[0019] In the document identification system proposed according to the first aspect of the utility model, a light source controller for the light-emitting device is also provided in the box, and the light source controller and the light-emitting device have multiple working modes, wherein the multiple working modes are distinguished in terms of the triggering method of the light source controller and the light-emitting method of the light-emitting device.

[0020] In the document recognition system according to the utility model, by performing recognition of document information based on images using a machine vision recognition method, the workload of staff can be significantly reduced and a higher recognition accuracy can be achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other features of the present invention will become apparent with reference to the accompanying drawings, in which:

[0022] Figure 1 A perspective view of the document identification system according to the present invention is shown from the front, wherein a portion of the outer wall of the box is removed for clarity;

[0023] Figure 2 From the back, it shows Figure 1 A perspective view of a document identification system, wherein a portion of the exterior of the box is removed for clarity;

[0024] Figure 3 The document identification system according to the present invention is shown from the oblique rear, wherein the outer wall of the back of the box is removed;

[0025] Figure 4 The paper pressing mechanism of the document identification system according to the utility model is shown separately;

[0026] Figure 5 An implementation of the document identification method according to the utility model is schematically shown. DETAILED DESCRIPTION

[0027] It is easy to understand that according to the technical solution of the utility model, without changing the essential spirit of the utility model, a person skilled in the art can propose a variety of interchangeable structural modes and implementation modes. Therefore, the following specific implementation modes and drawings are only exemplary descriptions of the technical solution of the utility model, and should not be regarded as the entirety of the utility model or as a limitation or restriction to the technical solution of the utility model.

[0028] The directional terms such as up, down, left, right, front, back, front, back, top, bottom, etc. mentioned or may be mentioned in this specification are defined relative to the structures shown in the drawings. They are relative concepts and may change accordingly according to their different positions and different usage states. Therefore, these or other directional terms should not be interpreted as restrictive terms. In addition, the terms "first", "second", "third", etc. or similar expressions are only used for description and distinction purposes, and cannot be understood as indicating or implying the relative importance of the corresponding components.

[0029] Before describing the document identification system and document identification method in detail according to the attached drawings, it should be noted that the document refers to the information about the configuration of the whole vehicle (such as the information of the assembled parts), the information related to the identity of the vehicle (such as the VIN code) and other relevant information specific to the vehicle during the whole vehicle manufacturing process, that is, the information to be identified mentioned in this article. Here, the document (which can also be called the whole vehicle document) can be but is not limited to a certificate of conformity, a certificate of conformity, an environmental protection list and a CCC (China Compulsory Certification) label.

[0030] refer to Figure 1 , Figure 2 and Figure 3 , which respectively present an embodiment of the document recognition system according to the utility model from different angles, and the document recognition system as a whole includes an image acquisition device 100 and a control device (not shown), wherein the image acquisition device is used to acquire an image of a document, and the control device is used to recognize document information based on the acquired image. Here, the control device (which is not shown in the drawings) can be an external computer device and can simultaneously undertake the document information recognition task and the control task of the image acquisition device, wherein the configuration of the control device will be described in detail below in conjunction with the document recognition method according to the utility model.

[0031] In summary, the document recognition system according to the present invention uses a visual recognition method based on an image, especially OCR technology (Optical Character Recognition) to recognize document information, thereby replacing the labor-intensive visual method and achieving higher recognition accuracy.

[0032] Specifically, a paper pressing mechanism 120, a camera 130, and a light source mechanism are arranged in the housing 110 of the image acquisition device, wherein the document to be identified is flattened and held on the paper pressing mechanism 120; the camera is positioned toward the paper pressing mechanism 120 to collect the image of the document located there; and the light source mechanism is used to provide a lighting environment for shooting for the camera 130. According to the utility model, the light source mechanism includes a light emitting device (not visible in the drawings) and a light shield 140 (wherein, for the sake of clarity, not all light shields are provided with reference numerals), wherein the light emitting device is fixed inside the light shield 140, specifically on the inner wall of the light shield, and the light shield 140 covers the paper pressing mechanism 120 on the outside, thereby providing a so-called dark box for camera shooting.

[0033] The box 110 is, for example, a tooling box, and is equipped with a Forma wheel at the bottom for the convenience of daily maintenance. In addition, the box fulcrum adopts adjustable anchor rollers and is combined with a level reference instrument optionally arranged inside the box to achieve accurate positioning and optimal horizontality. In the box 110, especially in its lower part, there is also a distribution box 150 for centralized power distribution to the image acquisition device, which can be fixed on the side wall of the box 110 and can also be equipped with an independent metal box for safety reasons. In a specific embodiment, in order to avoid overload, the distribution box 150 has redundant distribution power and a port for access to the production network.

[0034] In some embodiments, a plurality of heat dissipation slots 160 are formed on the side wall of the box 110 by cutting technology to dissipate heat when the equipment in the box is working. In addition, considering the possible influence of ambient temperature, humidity and dust during the overall operation, a fan 170, especially a silent fan, is also provided on the box to timely ventilate and dissipate heat inside the machine. Furthermore, a filter can be added at the fan 170 to ensure the cleanliness of the overall working environment and to avoid the influence of dust on the identification of document information as much as possible.

[0035] For the image-based document information recognition process, information capture and recognition are closely related to the image quality, and one of the main influencing factors is the camera configuration. In the document recognition system, in order to ensure accurate and clear recognition of the information on the paper, the camera used can be but not limited to a high-definition camera or a sheet-fed scanner. Among them, a high-definition camera with about 20 million pixels is more advantageous in terms of resolution, contrast, depth of field, etc., which is conducive to algorithm processing and has better stability.

[0036] In machine vision, another important factor for image quality and recognition accuracy is the lighting environment in which the camera is located, in which the features of the target object need to be highlighted while ignoring its background, and interference caused by external or self-reflection should be avoided. With this in mind, according to the utility model, the light source mechanism adopts a dome-type light source, which can achieve an effect similar to diffuse reflection of natural light, with uniform illumination and sufficient brightness, and thus can achieve a better imaging effect. Specifically, the dome-type light source mechanism includes a dome-type light shield 140, and the light-emitting device is fixed to the inner wall of the light shield 140, specifically to the side wall inside it, by means of a slide rail mechanism 141, and the light-emitting device can be pushed into or pulled out of the light shield by means of the slide rail mechanism. This push-pull fixing method facilitates the maintenance and replacement of the light-emitting device.

[0037] In some specific embodiments, the slide rail mechanism 141 is mounted on the left and right sides of the light shield and light emitting devices, especially LED rows, are accommodated in the receiving grooves of the slide rails, wherein the dome-type light shield can make the lighting effect softer, which is more conducive to printing, color recognition detection, engraving, damage and dirt detection, etc. The slide rail mechanism 141 can be composed of an outer fixed rail, an intermediate fixed rail, a retainer and an inner fixed rail, and rolling steel balls are distributed on each rail to make the operation smooth and prevent shaking during the push and pull process.

[0038] In some other embodiments, a light source controller 180 (or a voltage-stabilized light source controller) is also provided at the lower part of the housing 110, which communicates with each camera and is used to adjust the working state of the light-emitting device assigned to the camera. The light source controller and the corresponding light-emitting device have multiple working modes. Specifically, in the first working mode, it is set to a software-triggered constant light mode, in which button triggering and communication triggering are feasible. The "trigger" mentioned here refers to the triggering method of the light source controller by an external trigger. In the first working mode, the light-emitting device is connected to the digital light source controller and is always on, and the illumination brightness is adjusted by pulse width modulation. In the second working mode, it is set to a software-triggered flash mode, and button triggering and communication triggering are feasible, so as to control the light-emitting device to be turned on and off instantly by simulating the contact with an external hardware trigger signal by software. In the third working mode, it is set to an externally triggered flash mode, which is triggered by a rising edge and can only be triggered by an input port. Specifically, in the third working mode, the illumination intensity can only be triggered by an external signal, which belongs to an external signal rising edge trigger. In the fourth working mode, it is set to an externally triggered flash mode, which is triggered by a falling edge and can only be triggered by an input port. In the fourth working mode, the lighting intensity can only be triggered by an external signal, which belongs to the external signal falling edge triggering.

[0039] Combine the following Figures 1 to 3 To illustrate the relationship between the light source mechanism and the camera in terms of structure and arrangement, two documents can be placed on each paper pressing mechanism 120 (which will be described in detail below), such as a list of environmental information accompanying a light gasoline vehicle, a vehicle conformity certificate, a certificate of inspection and a CCC label, and corresponding cameras are provided for the front and back of these four documents. Specifically, two cameras are provided in the upper left part of the box 110, which are respectively used to capture single-sided images of the documents to be aligned, and the two cameras use a light shield located below them; two cameras are provided in the upper right part of the box 110 and use a light shield located below them (see Figure 2 ); Two cameras are provided in the lower left part of the box 110, which are used to collect single-sided images of the aligned documents, and the two cameras use the light shield located above them (see Figure 1); Two cameras are arranged in the lower right corner of the box and are illuminated by a light shield located above them (see Figure 1 ).

[0040] It should be noted that the number and position of the cameras are determined according to the documents to be identified. For example, it is feasible to have a camera only on one side and to capture the image of the opposite side of the document by turning it over. It is also feasible that the arrangement of the cameras matches the document to be identified, and if only one side of the document needs to be identified, a unique camera for that side is provided. In addition, it is also feasible to capture images of documents at different positions by moving a small number of cameras, which can be achieved by means of a movable slide.

[0041] The camera 130 may have an aperture locking bolt, an aperture value mark point, and a focal length locking bolt, wherein the aperture and focal length may be adjusted according to actual needs. After the aperture mark point is adjusted according to the light and the focal length is adjusted according to the actual imaging, the aperture fixing bolt and the focal length fixing bolt are used to fix them respectively.

[0042] Considering that the printed documents may bend or curl to varying degrees and the curvature value is irregular, which may cause image deformation, a dedicated paper pressing mechanism 120 is provided according to the utility model. Figure 4 The paper pressing mechanism comprises an upper light-transmitting plate 121 and a lower light-transmitting plate 122 which are placed flat against each other, wherein the upper light-transmitting plate 121 can rotate around its edge fixed to the lower light-transmitting plate 122 by means of a rotating bearing 123, so that the paper pressing mechanism 120 can be switched between an open state and a pressed state. In the open state, the document can be placed between the upper light-transmitting plate 121 and the lower light-transmitting plate 122, while in the pressed state, the document is pressed between the upper light-transmitting plate 121 and the lower light-transmitting plate 122. Here, the rotating bearing 123 adopts a high-precision optical axis and lengthens the rotating shaft to increase the force-bearing area and stability and reduce maintenance. In the open state, the angle between the upper light-transmitting plate 121 and the lower light-transmitting plate 122 can be up to about 11.83°, and the lower edge of the document can be, for example, 130 mm away from the front side of the box. In this open state, the staff can push the document into the paper pressing mechanism 120 or pull it out from the operating table 111. The operating table is made of stainless steel, for example, and is detachably fixed to the box body 110, for example, by a combination of four-corner bolts and drilled holes of matching shape.

[0043] Here, the upper and lower light-transmitting plates can be light-transmitting frosted glass, which can avoid specular reflection while flattening the paper, thereby ensuring soft diffuse reflection and an ideal lighting environment. The upper light-transmitting frosted glass is fixed by main beam screws, and the lower light-transmitting frosted glass is fixed to the bottom plate at the center. Two separate card slots for placing paper are provided on the bottom plate 124, and each card slot is surrounded by double-sided enclosures on one side and the top. In addition, a guide portion 125 extending outward to the edge is vertically provided on the bottom plate 124 to provide document guidance for operators.

[0044] In some embodiments, the paper pressing mechanism 120 also includes a handle mechanism 126, which is fixed to the upper light-transmitting plate 121 and extends out of the box body 110 with an operating handle 127, wherein a snap-fit ​​portion is also provided on the operating handle 127, which can be fixed to the outer wall of the box body 110 in the above-mentioned opened state.

[0045] In addition, a document recognition method that can be performed by the document recognition system according to any one or more of the above embodiments is also proposed. Figure 5 According to one embodiment, the document identification method includes the following steps:

[0046] S100: Acquire an image of the document captured by a camera;

[0047] S200: capturing a local image at a specific area in the image, and preprocessing the local image, wherein the specific area is an area containing information to be identified;

[0048] S300: performing character recognition on the pre-processed partial image, and generating a post-processing instruction accordingly according to the recognition result.

[0049] The preprocessing includes one or more of the following operations: grayscale processing, binarization processing, noise reduction processing, tilt correction processing and text segmentation processing.

[0050] Specifically, during the grayscale processing, the collected color partial image is converted into a grayscale image, wherein the grayscale value ranges from 0 to 255.

[0051] In the binarization process, the image data is divided into a pixel group smaller than the preset threshold and a pixel group larger than the preset threshold according to a preset threshold, wherein the preset threshold can be a fixed threshold or an adaptive threshold. Here, the binarization process can involve a bimodal method, a P parameter method, an iterative method, an OTSU method, and the like.

[0052] In the process of noise reduction, it may involve mean filter, adaptive Wiener filter, median filter, morphological noise filter, wavelet denoising, etc.

[0053] The tilt correction process is used to eliminate the image tilt caused by non-horizontal shooting, and Hough transform may be involved. The image is expanded to connect the discontinuous text into a straight line for easy straight line detection. After calculating the angle of the straight line, the tilted image is corrected to a horizontal position using a rotation algorithm.

[0054] For documents, the text segmentation process includes two steps: line segmentation and character segmentation. When performing line segmentation, the tilt-corrected text is projected onto the Y-axis, and all values ​​are accumulated to obtain a histogram on the Y-axis, where the bottom of the histogram is the background and the peak is the area where the foreground (text) is located. Character segmentation is similar to line segmentation. When performing character segmentation, each line of text is projected onto the X-axis. Here, when segmenting characters, the overlap between characters is taken into account, and a threshold for the character width is pre-set. If the projection of the cut character exceeds the threshold to a certain extent, it is determined that the cut character is actually two characters; if it is less than the threshold to a certain extent, the gap generated during the projection is ignored and the characters on the left and right of the gap are combined into one character for recognition.

[0055] After the above preprocessing, the segmented characters need to be recognized. After feature extraction of the text image, the character information in the document is recognized by using a classifier of a trained optical character recognition model to classify the characters.

[0056] In some embodiments, the document identification method further comprises the following steps:

[0057] The placement posture of the document in the paper pressing mechanism is detected, wherein if the document is determined to be in place, a trigger signal for collecting an image for identification is issued, otherwise a first warning signal is issued, and the placement posture of the document is adjusted according to the first warning signal. The detection of the placement posture of the document can be realized by a control device, wherein the determination is made based on the relative position of the document relative to the paper pressing mechanism or its card slot in the image captured by the camera. Alternatively, the detection of the placement posture of the document can be performed manually with the help of a special monitor, wherein the monitor is arranged in the box and is used to display the placement posture of the document relative to the paper pressing mechanism in real time.

[0058] In some other embodiments of the document identification method, the above-mentioned generating a post-processing instruction according to the identification result also includes the following steps:

[0059] If the character recognition determines that there is a longitudinal anomaly or a lateral anomaly, a second warning signal is issued, wherein the longitudinal anomaly is that the information to be identified does not match the stored standard information, and the lateral anomaly is that multiple pieces of information to be identified of the same type do not match each other;

[0060] Otherwise, a data storage signal is issued, according to which the identification image and the identified document information are stored as a group.

[0061] This involves identifying whether the information in a single document is correct or not and identifying the consistency of information between different documents. If it is determined that there is an error in the document information, the staff can be reminded to check. If the document information is determined to be correct, the relevant data will be stored in the cloud or database in a matching manner for subsequent accurate tracing. It is also feasible to present the recognition result to the staff, for example, by displaying it on an optional display of the document recognition system according to the utility model.

[0062] Taking the document recognition system shown in the attached figure as an example, the document recognition method implemented by it is explained. After four documents are placed and pressed in the paper pressing mechanism, the visual recognition system is started to collect the document image; then, the frame number, issuance date, etc. in the document image are captured and captured as a partial image; the partial image is preprocessed and the character features are recognized with the help of the optical character recognition model stored in the control device. If the document information is determined to be correct, the relevant data is transmitted to the downstream interface and stored, otherwise an alarm is issued. Here, not only the character information in a single document is recognized, but also the consistency of the same character information in multiple documents is recognized.

[0063] Here, the document recognition method according to the present invention particularly results in the above-mentioned features and advantages of the document recognition system according to the present invention, and reference can be made to the explanations thereof accordingly.

[0064] It should be understood that all the above preferred embodiments are illustrative rather than restrictive, and various modifications or variations made by those skilled in the art to the specific embodiments described above under the concept of the present invention should be within the legal protection scope of the present invention.

Claims

1. A document recognition system, characterized in that: It includes an image acquisition device and a control device, the control device is configured to recognize document information based on the image, and the image acquisition device includes: Box; A paper pressing mechanism, which is fixed in the box and is used to flatten and hold the document therein; a camera, which is fixed in the box toward the paper pressing mechanism to obtain an image of the document at the paper pressing mechanism, and the image is transmitted to the control device; The light source mechanism comprises a light emitting device and a light shield, wherein the light emitting device is fixed inside the light shield, and the light shield covers a paper pressing mechanism on the outside and is used to provide a lighting environment for shooting.

2. The document recognition system according to claim 1, characterized in that: The light shield is dome-shaped, and a light emitting device is fixed on the inner wall of the light shield by means of a slide rail mechanism, and the light emitting device is pushed into the light shield or pulled out from the light shield by means of the slide rail mechanism.

3. The document recognition system according to claim 1, characterized in that: The paper pressing mechanism includes an upper light-transmitting plate and a lower light-transmitting plate placed flat against each other, wherein the upper light-transmitting plate can rotate around its edge fixed to the lower light-transmitting plate by means of a rotating bearing, so that the paper pressing mechanism can be switched between an open state and a pressed state. In the open state, the document can be placed between the upper light-transmitting plate and the lower light-transmitting plate, and in the pressed state, the document is pressed tightly between the upper light-transmitting plate and the lower light-transmitting plate.

4. The document recognition system according to claim 3, characterized in that: The upper light-transmitting plate or the lower light-transmitting plate is light-transmitting frosted glass.

5. The document recognition system according to claim 3, characterized in that: The paper pressing mechanism also includes a handle mechanism, which is fixed to the upper light-transmitting plate and extends out of the box body with an operating handle, wherein a buckle portion is provided on the operating handle, and the buckle portion can be fixed to the outer wall of the box body in the opened state.

6. The document recognition system according to claim 3, characterized in that: Cameras are arranged above the upper light-transmitting plate and below the lower light-transmitting plate to capture images of the front and back of the document.

7. The document recognition system according to claim 1, characterized in that: The box body is provided with a heat dissipation slot.

8. The document recognition system according to claim 1, characterized in that: A fan is fixed on the box to perform convection heat dissipation inside the box.

9. The document recognition system according to claim 8, characterized in that: A filter is also provided at the fan to filter the air entering the box.

10. The document recognition system according to claim 1, characterized in that: A light source controller for the light emitting device is also provided in the housing. The light source controller and the light emitting device have a plurality of operating modes, wherein the plurality of operating modes are distinguished in terms of a triggering mode of the light source controller and a light emitting mode of the light emitting device.

Images