Positioning method, control method and device for coke oven four cars based on computer vision

Abstract

This invention provides a computer vision-based positioning method, control method, and apparatus for the four main locomotives of a coke oven. The method includes controlling a camera mounted on the locomotive to capture images in real time; extracting a first image captured by the camera when the furnace number marker is within the camera's field of view; comparing the first image with various second images in a pre-stored image library to determine the furnace number of the locomotive; and determining the offset between the camera and the corresponding furnace number marker, wherein each second image in the image library is an image obtained by capturing images of each furnace number marker; and determining positioning information based on the furnace number and the offset. Implementing this invention achieves a balance between positioning accuracy and cost.

Description

Technical Field

[0001] This invention belongs to the field of computer industrial application technology, specifically relating to a computer vision-based positioning method, control method, and device for the four main coke oven cars. Background Technology

[0002] The goal of the four major coke oven locomotive automation systems is to achieve coke pushing interlocking and coordinated operation of the four locomotives through precise measurement of their respective positions and reliable data communication, as well as automatic driving, positioning, and operation under production control. This will enable automated production operation, computerized production management, and ultimately, unmanned driving of the locomotives.

[0003] Locomotive positioning technology is a key technology for the automation of coke oven locomotives and a core issue in achieving unmanned operation. Current positioning technologies both domestically and internationally mainly include rotary encoder position detection, barcode reader-based position detection, infrared positioning technology, coded cable position detection, barcode position detection, and electromagnetic induction technology. The four-car positioning technology currently used in coking plants has the following drawbacks:

[0004] 1) Rotary encoder position detection technology: long-term operation results in large cumulative errors and low positioning accuracy;

[0005] 2) Code reader encoding position detection and infrared positioning technology: can only locate a single point, has low reliability, high failure rate, and high maintenance cost;

[0006] 3) Encoded cable position detection: The positioning accuracy is up to ±5mm, which is not high. Once damaged, the entire encoded cable needs to be replaced, resulting in high maintenance costs. Summary of the Invention

[0007] To overcome at least one technical problem in the prior art, the present invention provides a computer vision-based positioning method, control method and device for the four main coke oven cars, which can balance accuracy and cost.

[0008] The technical solution of this invention is: a computer vision-based positioning method for the four main coke oven carts, comprising:

[0009] The camera installed on the locomotive is controlled to capture images in real time, and the first image captured by the camera when the furnace number marker is within the camera's field of view is extracted;

[0010] The first image is compared with each second image in the pre-stored image library to identify the furnace number of the locomotive, and the offset between the camera and the corresponding furnace number marker is extracted, wherein each second image in the image library is an image obtained by taking pictures of each furnace number marker;

[0011] The positioning information is determined based on the furnace number and the offset.

[0012] Furthermore, before comparing the first image with each of the second images in the pre-stored image library, the method further includes:

[0013] The first image is preprocessed, and the preprocessed first image is then subjected to edge extraction processing. The edge-extracted first image is then compared with the second image.

[0014] Furthermore, the preprocessing includes at least one of noise reduction processing and image enhancement processing, and the edge extraction processing includes at least one of binarization processing and grayscale value processing.

[0015] Furthermore, the furnace number identification object includes a ruler, color card, or digital furnace number identification plate.

[0016] Further, comparing the first image with each second image in a pre-stored image library to identify the furnace number of the locomotive includes:

[0017] Based on the comparison results of comparing the first image with each second image in the pre-stored image library, a second image with a similarity exceeding a preset value to the first image is determined;

[0018] The furnace number corresponding to the second image is taken as the furnace number where the locomotive is located.

[0019] Furthermore, the positioning information includes the furnace number and the offset, or the positioning information includes determining the relative position between the locomotive and the furnace number based on the furnace number and the offset.

[0020] One technical solution of the control method of the present invention is to send the positioning information to the central control module, and the central control module performs four-vehicle interlock control.

[0021] The technical solution of a positioning device of the present invention includes:

[0022] The extraction module is used to control the camera installed on the locomotive to capture images in real time and extract the first image captured by the camera when the furnace number marker is within the camera's shooting range;

[0023] The first determining module is used to compare the first image with each second image in a pre-stored image library to identify the furnace number where the locomotive is located, and to extract the offset between the camera and the corresponding furnace number marker, wherein each second image in the image library is an image obtained by taking pictures of each furnace number marker;

[0024] The second determining module is used to determine the positioning information based on the furnace number and the offset.

[0025] Furthermore, the device further includes a processing module, configured to preprocess the first image before comparing it with each second image in a pre-stored image library, and to perform edge extraction processing on the preprocessed first image.

[0026] The technical solution of a control device of the present invention includes a sending module and a positioning device; the sending module is used to send positioning information determined by a second determining module to a PLC central control module; the central control module is used to perform four-vehicle interlocking control based on the positioning information.

[0027] The present invention has the following technical effects:

[0028] 1) The positioning method of the present invention has higher positioning accuracy, which enables the four-car interlocking control to ensure the smooth and safe operation of the four cars, avoiding accidents caused by positioning problems of the four cars during production. This increases the service life of the coke oven locomotive and the coke oven body, ensures that the operation is strictly carried out according to the process, and greatly improves the quality of coke. This also improves the degree of automation in production, increases production efficiency, and reduces the labor intensity of operators, bringing great economic benefits overall.

[0029] 2) This invention utilizes computer vision, which can quickly acquire large amounts of information, is low-cost, and easy to automate. It has been successfully applied in industrial production, significantly improving the level of automation and the quality and reliability of products. The hardware configuration for computer vision is simple: an industrial computer, a high-speed industrial camera, and a furnace number marker are sufficient to accurately locate the target furnace number, resulting in low hardware costs.

[0030] 3) This invention has high positioning accuracy and low equipment maintenance costs. The application of computer vision positioning technology in the positioning of four vehicles in the coking industry is a world first. Attached Figure Description

[0031] Figure 1 This is a flowchart of the present invention. Detailed Implementation

[0032] The main equipment of a coke oven includes the coke oven heating exchange equipment (exchange system), as well as the coke pusher car, coke quencher car, coke quenching car, and coal charging car, collectively referred to as the "four cars and one machine," the latter often called the "four major cars." Coke oven production operations are completed through the coordinated operation of these various machines. With the gradual improvement of coke oven machinery, the coordination between the four major cars, their communication, address and furnace number alignment, coke pusher action interlocking, coal charging action interlocking, and ultimately, centralized computer control—all these automation technology solutions are of great significance for improving the production management level, product quality, and economic benefits of coking plants. At the same time, the implementation of the four-car interlocking system ensures production safety and strict adherence to planned production schedules.

[0033] The key technology for achieving four-car interlocking is the precise positioning of the furnace number of each of the four cars. The application of computer vision positioning technology is an effective method for achieving precise positioning of the target furnace number of each of the four cars.

[0034] Example 1:

[0035] The computer vision-based localization methods for the four main coke oven carts include:

[0036] The camera installed on the locomotive is controlled to capture images in real time, and the first image captured by the camera when the furnace number marker is within the camera's field of view is extracted;

[0037] The first image is compared with each second image in a pre-stored image library to identify the furnace number of the locomotive, and the offset between the camera and the corresponding furnace number marker is extracted, wherein each second image in the image library is an image obtained by taking pictures of each furnace number marker; positioning information is determined based on the furnace number and the offset.

[0038] For example, the offset between the camera center coordinates and the corresponding furnace number marker center coordinates is calculated and used as the offset between the camera and the corresponding furnace number marker. The offset between the camera center coordinates and the corresponding furnace number marker center coordinates can be obtained using existing methods for calculating the offset between the camera center coordinates and the center coordinates of objects in the captured image.

[0039] Furthermore, before comparing the first image with each of the second images in the pre-stored image library, the method further includes:

[0040] The first image is preprocessed, and the preprocessed first image is then subjected to edge extraction processing. The edge-extracted first image is then compared with the second image.

[0041] Furthermore, the above preprocessing includes at least one of noise reduction processing and image enhancement processing, and the edge extraction processing includes at least one of binarization processing and grayscale value processing.

[0042] Furthermore, the aforementioned furnace number identifiers include rulers, color cards, or numerical furnace number identifier plates.

[0043] Furthermore, the comparison of the first image with various second images in a pre-stored image library to identify the furnace number of the locomotive includes:

[0044] Based on the comparison results of comparing the first image with each second image in the pre-stored image library, a second image with a similarity exceeding a preset value to the first image is determined;

[0045] The furnace number corresponding to the second image is taken as the furnace number where the locomotive is located.

[0046] Furthermore, the aforementioned positioning information includes the furnace number and the offset, or the positioning information includes determining the relative position between the locomotive and the furnace number based on the furnace number and the offset.

[0047] One technical solution of the control method of the present invention is to send the positioning information to the central control module, and the central control module performs four-vehicle interlock control.

[0048] The technical solution of a positioning device of the present invention includes:

[0049] The extraction module is used to control the camera installed on the locomotive to capture images in real time and extract the first image captured by the camera when the furnace number marker is within the camera's shooting range;

[0050] The first determining module is used to compare the first image with each second image in a pre-stored image library to identify the furnace number where the locomotive is located, and to extract the offset between the camera and the corresponding furnace number marker, wherein each second image in the image library is an image obtained by taking pictures of each furnace number marker;

[0051] The second determining module is used to determine positioning information (which may be the furnace number and the offset itself) based on the furnace number and the offset. Further, the device also includes a processing module, used to preprocess the first image before comparing it with each second image in a pre-stored image library, and to perform edge extraction processing on the preprocessed first image.

[0052] A control device according to the present invention includes a sending module and a positioning device; the sending module is used to send positioning information determined by a second determining module to a PLC central control module; the central control module is used to perform four-vehicle interlocking control based on the positioning information.

[0053] Example 2:

[0054] like Figure 1 As shown, before the system identifies the furnace number, images of all furnace number identification plates are captured to generate an image library. During system operation, the camera is turned on and captures images in real time. When the camera vehicle moves and a furnace number identification plate enters the camera's range, the camera performs image recognition and comparison on the identification plate. Halcon vision graphics processing software is used to analyze the target furnace number in real time and to analyze the offset between the camera center coordinates and the furnace number identification plate center coordinates. With the furnace number and the offset, the relative position of the locomotive with respect to the center of the furnace number identification plate can be determined. After software analysis and processing, the industrial control computer transmits the identified furnace number and coordinate offset to the central control PLC via wireless communication. The PLC then controls the movement and performs four-vehicle interlocking control based on the positioning position.

[0055] System Composition: The computer vision positioning system for the four main coke oven cars consists of an industrial control computer, a high-resolution variable focal length industrial high-speed camera, an image acquisition card, a furnace number identification plate, a stable light source, an AP+CPE wireless network, and a central control PLC.

[0056] Image recognition software is the core and key technology of the entire computer vision positioning system; the quality of its algorithm directly affects the system's operating speed and recognition accuracy. Image recognition software mainly consists of three parts: image preprocessing, edge extraction, and image recognition. The software flowchart for image recognition software is as follows:

[0057] (1) Image preprocessing:

[0058] Number plate images captured by industrial cameras inevitably generate some noise during transmission and conversion. Therefore, image preprocessing, including noise reduction and image enhancement, is necessary before image processing. Noise poses many difficulties for image processing and directly affects subsequent image processing tasks such as image segmentation, edge extraction, and image recognition. Therefore, real-time acquired number plate images require filtering.

[0059] (2) Edge extraction:

[0060] In order to separate the furnace number identification plate image from the complex background in image recognition processing, the obtained part edge image is binarized and grayscale value processed.

[0061] (3) Image recognition:

[0062] When the locomotive is moving, the system compares real-time captured images with pre-stored images of the identification plates to analyze the offset of the license plate numbers and the camera center from the furnace number identification plate. The data is then transmitted wirelessly to the central control PLC for interlocking control.

[0063] Regarding the selection of furnace number identification plates in the technical solution, digital identification plates are only one type of identification object, and furnace number identification plates can also be replaced by other means such as rulers or color cards.

[0064] In summary, this invention offers fast positioning speed and can operate continuously, up to 24 hours a day. In coking plant production, traditional methods still rely heavily on operators visually identifying furnace numbers and aligning the locomotive. Prolonged operation leads to eye fatigue and emotional instability, potentially causing production accidents. Slight differences in how operators judge the center of the target furnace number also exist, requiring highly experienced operators. During periods of high production volume, rapid manual positioning is unreliable. Computer vision positioning technology not only overcomes the shortcomings of traditional manual furnace number positioning methods but also leverages its advantages of speed and accuracy.

Claims

1. A computer vision-based positioning method for the four main coke oven carts, characterized in that: include: The camera installed on the locomotive is controlled to capture images in real time, and the first image captured by the camera when the furnace number marker is within the camera's field of view is extracted; The first image is compared with each second image in the pre-stored image library to identify the furnace number of the locomotive, and the offset between the camera and the corresponding furnace number marker is extracted, wherein each second image in the image library is an image obtained by taking pictures of each furnace number marker; The positioning information is determined based on the furnace number and the offset. Before comparing the first image with each of the second images in a pre-stored image library, the method further includes: The first image is preprocessed, and the preprocessed first image is then subjected to edge extraction processing. The edge-extracted first image is then compared with the second image. The preprocessing includes at least one of noise reduction processing and image enhancement processing, and the edge extraction processing includes at least one of binarization processing and grayscale value processing.

2. The method according to claim 1, characterized in that, The furnace number identification includes a ruler, color card, or numerical furnace number identification plate.

3. The method according to claim 1, characterized in that, The step of comparing the first image with each second image in a pre-stored image library to identify the furnace number of the locomotive includes: Based on the comparison results of comparing the first image with each second image in the pre-stored image library, a second image with a similarity exceeding a preset value to the first image is determined; The furnace number corresponding to the second image is taken as the furnace number where the locomotive is located.

4. The method according to claim 1, characterized in that, The positioning information includes the furnace number and the offset, or the positioning information includes determining the relative position between the locomotive and the furnace number based on the furnace number and the offset.

5. A control method, characterized in that, The positioning information determined by the method according to any one of claims 1 to 4 is sent to the central control module, which then performs four-vehicle interlocking control.

6. A positioning device, characterized in that, The method described in claims 1 to 5 includes: The extraction module is used to control the camera installed on the locomotive to capture images in real time and extract the first image captured by the camera when the furnace number marker is within the camera's shooting range; The first determining module is used to compare the first image with each second image in a pre-stored image library to identify the furnace number where the locomotive is located, and to extract the offset between the camera and the corresponding furnace number marker, wherein each second image in the image library is an image obtained by taking pictures of each furnace number marker; The second determining module is used to determine positioning information based on the furnace number and the offset; The positioning device further includes: The processing module is used to preprocess the first image before comparing it with each second image in a pre-stored image library, and to perform edge extraction processing on the preprocessed first image.

7. A control device, characterized in that, Includes a sending module and a positioning device according to claim 6; The sending module is used to send the positioning information determined by the second determining module to the PLC central control module; The central control module is used to perform four-vehicle interlocking control based on the positioning information.

Images