JPH0423889A - Apparatus for recognizing state of throat of coke oven - Google Patents

Abstract

PURPOSE:To clearly distinguish the luminance difference between a throat and a furnace wall by attaching an optical filter having a prescribed characteristics to an image-pickup device of an image-recognition apparatus to recognize the state of the throat of a coke oven. CONSTITUTION:Image pickup 1, 2 having the image-pickup range including the throat 8 are supported in a state freely movable along a coke oven 13. Image of one frame of the image-pickup device is stored in a memory. The state of the throat is recognized by gaining access to the memory, reading the pixel of an arbitrary address, and processing the image using the luminance difference between pixels. An optical filter 9 having prescribed characteristics is attached to the image-pickup device to clearly distinguish the luminance difference between the throat 8 and the furnace wall 7.

Description

[Detailed Description of the Invention] [Industrial Field of Application] The present invention is a method for controlling the condition and position of the coke oven opening.
This invention relates to a reactor mouth situation recognition device that performs processing and identification, and relates to an improvement of the reactor mouth situation gait disclosed in Japanese Patent Application Laid-Open No. 157577/1983.

[Prior Art] According to a furnace mouth situation recognition device disclosed in JP-A No. 61-157577, the furnace mouth situation m1lil is determined by determining the brightness difference between the furnace mouth and walls using an imaging device, and it is possible to detect coke oven equipment. It is possible to understand the thermal expansion peculiar to coke oven equipment and changes in oven conditions due to aging without having to place workers on-site.
This enables smooth operation of coke oven equipment.

[Problem 1 to be solved by this invention: The furnace mouth situation according to the above-mentioned Japanese Patent Application Laid-open No. 157577/1983! ! The detection device is configured to determine the condition of the furnace by the difference in brightness between the inside of the furnace and the furnace wall.In coke ovens that are normally installed outdoors, during nighttime operation, the device detects the condition of the furnace by the difference in brightness. The mouth can be easily identified. However, for example, when exposed to direct sunlight,
In some cases, the difference in brightness between the furnace mouth and the furnace wall canceled out, or the brightness at the wall was higher than inside the furnace, making it difficult to distinguish the furnace status based on the difference in brightness. . One way to remove the effects of direct sunlight during the day is to cover the image detection field of view, but this was not a practical solution as it required covering not only the furnace mouth but also the area around it.

Therefore, the present invention provides a furnace mouth situation recognition device that is relatively inexpensive, easy to install, and can reliably grasp changes in the furnace mouth situation due to differences in brightness without being affected by external conditions such as direct sunlight. The purpose is to provide.

[Means for Solving the Problems] According to the coke oven oven mouth situation recognition device according to the present invention, there is provided an imaging device that is movably supported along the coke oven and uses the oven opening as an imaging area, and In an image recognition device that stores one frame image and accesses these memories to read pixels at arbitrary addresses and performs image processing based on the difference in pixel brightness, it is possible to recognize the reactor mouth situation. It is characterized by the fact that an optical filter with predetermined characteristics is used, which makes it possible to clearly identify differences in brightness at the furnace mouth and walls.

[Example] Hereinafter, the present invention will be explained using the illustrated example. 1st
The figure is a perspective view showing the configuration of the embodiment device, and FIG. 2 is a block diagram of the main parts thereof. As shown in FIGS. 1 and 2, two video cameras are installed in the vicinity of the furnace port 8 surrounded by the furnace frame 14 of the Kuuss class 3 in the vertical direction in which the furnace port 8 extends. 1°2 (only the upper one is shown in Figure 1). These video cameras 1 and 2 have their imaging centers aligned in the vertical direction, and their camera fields of view 12 are capable of imaging the upper part of the short day and the bottom of the reactor mouth, respectively, and are provided continuously in the lateral direction]- It is possible to move the mouse 13 along the furnace mouth 8 while maintaining the vertical relationship.

A filter 9 having predetermined optical characteristics is attached to these video cameras 1 , 2 , and the outputs of these video cameras 1 , 2 are further input to an image processing device 10 . The image processing device M1o includes video cameras 1 and 2, respectively.
A pair of video memories 3.4 are provided corresponding to the video cameras 1.2 and store one frame of images taken by the video cameras 1.2 as pixel data. moreover,
This image processing device 10 includes a pair of video memories 3 and 4.
An image recognition device M5 is provided which is optionally connected to the image switching device 6 via the image switching device 6. This image switching device 6 is capable of switching its input to one of the video memories 3.4 and outputting the contents to the image recognition device 5 according to a command from the image recognition device [5].

Further, the image recognition device @5 sends a signal to the image switcher 6 to switch to any one of the video memories 3.4, and
By specifying a pixel address in a frame of an image stored in each video memory 3.4, pixel data at an arbitrary address can be read. In the apparatus configured in this way, the video camera 1.2 is placed on the oven side of the coke oven working machine, and is placed vertically below the imaging center of the video cameras 1 and 2 while checking on the attached monitor 11.
The equipment can be adjusted and installed so that the imaging centers of the two are aligned, and the equipment can be placed away from the coke oven. The images captured by these two video cameras 1.2 yield two images, an image of the top of the furnace mouth and an image of the bottom of the furnace mouth, as shown in Figure 1. Similarly, due to the presence of red-hot coke 15 in the coke oven mouth 8, the inside of the oven mouth 8 is a bright area with high brightness, and the oven wall is a screen with a low brightness dark area.

Here, the video camera 1, which is a characteristic part of this embodiment,
To explain in detail about the filter 9 attached to the filter 2, the coke F113 is usually installed outdoors for operational reasons, and when it is not affected by external light such as at night, the brightness difference as described above occurs. (Light emission due to high temperature) to furnace wall 7
and the furnace mouth 8 can be clearly distinguished on the captured image. However, as illustrated in FIG. 1, when exposed to direct sunlight, the brightness characteristics of the furnace body 13 itself as shown in FIG. The brightness distribution ■ due to sunlight will be imaged as Ill. That is, in the example shown in (2) in FIG. 4, the brightness level of the furnace wall 7 becomes higher due to direct sunlight than the brightness of the furnace port 8 due to the red heat of the coke, and this is used as is to identify the furnace port. This makes it impossible to accurately determine the situation at the reactor mouth.

Here, as shown in Figure 5, the wavelength and radiant energy characteristics of sunlight are such that most of its energy is 50On-
The peak is concentrated in the area of pride. On the other hand, as shown in the spectral characteristics actually measured within the multiple furnace ports 8 shown in (2), O, and (c) of FIG. 6, the spectral characteristics within the coke furnace port 8 are 70 On I noticed that the energy was concentrated in the area. Therefore, in this embodiment, the difference in spectral characteristics between sunlight and the inside of the furnace port 8 is utilized to block most of the peak energy of the sunlight and allow most of the light energy inside the furnace port 8 to pass through. This problem was solved by adopting the optical filter 9.

Here, an optical filter having a characteristic of transmitting light of 65 Onm or more is adopted as the optical filter 9.
It is attached to video camera 1.2. By adopting this filter 9, the brightness distribution characteristics due to direct sunlight as shown in the fourth figure (■) are changed from the brightness distribution characteristics caused only by the heat generation inside the coke oven, as illustrated in the figure (■). Only the characteristics can be extracted, and if this is binarized using a predetermined threshold, the coke oven mouth 8 is determined to be bright, and the oven wall 7 is determined to be bright, as shown in Figure 4 0■. Brightness characteristics that allow both to be sufficiently distinguished are obtained.

In the apparatus with the above configuration, 1jll19B at the upper part of the furnace mouth
When performing the crawling, the image II device f5 switches the image switch 6 to the video memory 3 side so that an image of the upper part of the reactor mouth can be obtained, outputs a freeze signal to the video memory 3, and outputs a freeze signal to the video camera 1. The images of the upper part of the reactor mouth sent by mail are stored. The image is the I10 of the image recognition device.
When a pixel address corresponding to a pixel required for processing in the video memory 3 is output from a boat (not shown), the corresponding pixel data is outputted from the video memory 3 by sequentially shifting the pixel address of the image recognition device for one screen. image data is obtained. This pixel data has information on the darkness of light.

image:! The detection device 5 performs the above processing to obtain one screen worth of image 1-2, and performs processing for detecting the furnace mouth core. First, in order to detect the boundary a at the upper end of the furnace mouth, The vertical pixel addresses at the center of the screen are output sequentially from the top of the screen downwards.

Here's a picture in advance! ! ! The brightness of the furnace mouth is set in the identification W5, and while inputting pixel data, it is compared with the set brightness of the furnace mouth and searches for pixel data that has a brightness higher than the brightness of the furnace mouth, and when the corresponding pixel data is detected, it is detected. It starts integrating the number of times and continues to compare the pixel data above. When this integrated value is greater than or equal to the number of integrated values defined as inside the furnace mouth (set), pixel addresses with pixel data with a brightness higher than the first detected furnace brightness are assigned to the furnace wall 7 and the upper end of the furnace mouth 8. The destination pixel address is stored as the boundary.

Next, in order to detect the boundary between the left end of the furnace mouth and the right end of the furnace mouth, as shown in Figure 8, the pixel address in the horizontal direction is moved from the left end of the screen to the right end from several lines above the pixel address at the top end of the furnace mouth. The pixel data is sequentially output, and as the pixel data is input, it is compared with the set brightness at the furnace mouth, and pixel data having a brightness equal to or higher than the brightness at the furnace mouth is searched. If the corresponding pixel data is not detected even after searching horizontally to the center of the screen, pixel data in the horizontal direction one line below is searched.

While searching in this way, when the corresponding pixel data is detected, the number of detections starts to be accumulated, and the comparison of the pixel data in the right direction is continued. Here, when the integrated number is greater than or equal to the set integrated value inside the furnace mouth, the pixel address having pixel data with a brightness higher than the first detected furnace mouth brightness is set as the boundary between f1117 and the left end of the furnace mouth 8, and the pixel address is set as the same pixel address. Memorize b.

Next, in order to detect the right edge of the furnace mouth, we continue searching in the direction, and when the incoming pixel data has a brightness less than the furnace mouth brightness, we start accumulating the number of detections as pixel data of the furnace wall. Continue comparing directional pixel data.

When the integrated value is equal to or greater than the integrated number specified as the furnace wall, the pixel address with the pixel data whose brightness is higher than the last detected furnace mouth brightness is set as the boundary between the furnace mouth 8 and the furnace wall 7. The same pixel address C is stored. Then, based on the stored pixel address, the number of pixels at the furnace mouth 8 is determined and stored using the following equation (1).

Number of pixels at the furnace mouth = pixel address at the right end of the furnace mouth - pixel address at the left end of the furnace mouth
(1) The above process is performed until the bottom of the screen, and the number of pixels at the furnace mouth for each horizontal line is stored.

Then, the maximum number of pixels at the furnace mouth is found from among all or the number of pixels for each horizontal line stored, and the center of the upper part of the furnace mouth is calculated from the left end address of the furnace mouth and the right end address of the furnace mouth using the following formula (2). Obtain pixel address e.

Center pixel address at the top of the furnace mouth - pixel address at the right end of the furnace mouth - pixel address at the left end of the furnace mouth... (2) Furthermore, the pixel address d at the center of the screen is determined by the following formula (3)
The center deviation f of the upper part of the furnace mouth is calculated from the above and stored.

Center deviation at the top of the furnace mouth = (pixel address at the center of the screen - center pixel address at the top of the furnace mouth) Therefore, the image recognition device @5 switches the image switching device 6 to the video memory 4 side so that an image of the bottom of the furnace mouth 8 can be obtained. Then, a freeze signal is output to the video memory 4 to store image data for a new screen at the bottom of the furnace port 8, and the pixel address is output in the same way as described above, and the designated pixel data is input, and the image data of the new screen at the bottom of the furnace port 8 is stored. Performs center detection processing. Here, in order to detect the boundary of the lower end of the furnace mouth 8, as shown in Figure 9, the vertical pixel addresses of the center of the screen are sequentially output from the lower end of the screen upwards, and the pixel data is input, and the same as above. The pixel address b of the boundary between the furnace wall 7 and the lower end of the furnace mouth 8 is found and stored using the method.

Next, in order to detect the boundary between the left end of the furnace port 8 and the right end of the furnace port 8, the image collection address is sequentially output from the left end to the right end on the screen as shown in FIG. 10, and pixel data is input. After this,
Using the same method as above, find the pixel address b at the left end of the furnace mouth and the pixel address C at the right end of the furnace mouth. 1) Calculate and memorize the number of self-pixels using formula 1). Repeat the operation to find and store the number of pixels for each line.

Then, the maximum number of pixels at the furnace mouth is found from among all the stored pixel numbers, and from the pixel address b at the left end of the furnace mouth and the pixel address C at the right end of the furnace mouth, the center of the furnace mouth bottom is found using equation (2). The center deviation i of the bottom of the furnace mouth is determined from the pixel address h using equation (3).

In the case where the situation of the furnace mouth 8 or the wall 7 by identifying the brightness difference for each pixel address is used to recognize other furnace conditions,
For example, a case will be described in which the method is applied to checking the attachment and detachment of a furnace lid. In this case, after the furnace lid membrane is completed,
Furnace port 8 or furnace I! using the above-mentioned detection method! ! 7 Which boundary reaches the detection ratio, it is determined as the furnace lid membrane (between the furnace mouth), and conversely, after the furnace lid is completed, the same detection method is used to detect the furnace mouth 8 or between the furnace mouth 8 and the furnace I!
17 Which boundary cannot be detected, making it impossible to detect the
can be confirmed without contact in a short time. In addition, it goes without saying that the present invention can be similarly applied to the grasping and recognition of the status of the furnace described in Japanese Patent Application Laid-Open No. 61-157577.

Here, the optical filter 9 attached to the video camera 12, which is an imaging device, blocks the wavelength band of the main energy distribution of sunlight, but on the other hand, it blocks the wavelength band of the main energy distribution of light emitted by the heat generated by the coke in the furnace mouth. Since the filter passes through the band, once installed, it is necessary to remove the filter not only at night when there is no direct sunlight, but also during the day when the brightness of the wall tends to be high due to direct sunlight. It is possible to perform its original function without any problems.

[Effects of the Invention] In this embodiment, two video cameras 1.2 are used, but the configuration is not limited to this, and one camera can be used only for either the top or the bottom of the furnace mouth. It is also possible to take an image of the reactor and process it using the data obtained from this, or it can be structured to scan along the vertical direction of the furnace mouth with a single video camera.

According to this invention, the condition of the coke oven mouth m! The embodiment of the device is as described above, and the following effects can be achieved.

Even in an environment where the brightness of the furnace wall exceeds the luminance of the inside of the furnace itself, such as when it is exposed to direct sunlight, the furnace wall and furnace opening can be illuminated based on the luminous effect caused by the heat generated by the coke. Highly reliable hearth shape that can be accurately identified by brightness difference ii
! I can do I.

[Brief explanation of the drawing]

Fig. 1 is a perspective view showing the configuration of the embodiment, Fig. 2 is a block diagram of the embodiment, Fig. 3 is a schematic diagram showing the imaging state of the upper and bottom parts of the reactor mouth, and Figs. Figure (f) shows the brightness characteristics due to direct sunlight near the reactor mouth, the brightness characteristics after filter processing, and the characteristics after binarization processing at the threshold. Figure 5 shows the radiant energy and wavelength characteristics of sunlight. 6th diagram, 6th
Figures 6 and 6 are diagrams of actually measured spectral characteristics inside the furnace mouth, Figures 7 and 8 are diagrams of the upper part of the furnace mouth recognition process, and Figures 9 and 10.
The figure is an example of the furnace mouth bottom recognition process. Care, 701090. Furnace wall, 8... Furnace mouth, 9...
... Filter 10 ... Image processing device, 1
1... Monitor 12... Camera field of view, 13... Coke oven, 14... Furnace frame, 15... Red-hot coke, 16. ... Direct sunlight, a ... Upper end of the furnace mouth, b ... Left end pixel address of the furnace mouth, C ... Right end pixel address of the furnace mouth, d - ... -・Pixel address at the center of the screen, f...
... Center deviation at the top of the furnace mouth, Q ... Pixel address at the bottom end of the furnace mouth, h ... Center pixel address at the bottom of the furnace mouth.

Claims (2)

[Claims] (1) Imaging devices arranged vertically with their horizontal imaging centers aligned and supported movably along the coke oven mouth, whose imaging area is the coke oven mouth; A coke machine equipped with a memory that stores a single frame image of the furnace mouth captured by an imaging device, and an image recognition device that accesses the memory, reads pixels at arbitrary addresses, and grasps the furnace mouth situation based on the brightness difference between the pixels. A coke oven furnace mouth situation recognition device, wherein the imaging device is provided with an optical filter having predetermined characteristics. (2) A coke oven furnace condition recognition device, wherein the optical filter has an optical property of transmitting light of 650 nm or more.