A waste furnace feeding monitoring system and method based on image processing

Abstract

The present invention relates to the technical field of image processing, and particularly relates to a waste charging furnace monitoring system and method based on image processing. The method includes: dividing the acquired infrared image after the waste is charged into the furnace to obtain a plurality of sub-regions; using the center of each sub-region as a seed point to perform superpixel segmentation on the infrared image to obtain a plurality of superpixel blocks; calculating the variance of the average temperature within all superpixel blocks; when the variance is greater than a threshold, it indicates that the temperature distribution in the furnace is uneven and an alarm is issued. That is, the solution of the present invention can accurately detect the temperature in the furnace.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing. More specifically, the present invention relates to a waste feeding furnace monitoring system and method based on image processing. Background Art

[0002] With the increasing generation of municipal domestic waste and industrial waste, traditional waste treatment methods such as landfilling not only occupy a large amount of land resources but also may cause problems such as groundwater pollution. Therefore, waste incineration technology has emerged as an important waste treatment means.

[0003] During the process of waste feeding into the furnace for incineration, it is necessary to capture infrared images in real time to monitor the temperature distribution inside the furnace, facilitating the timely detection of abnormal areas, preventing overheating and uneven temperature, and improving incineration efficiency and safety. At the same time, the captured infrared images also help to detect whether the equipment and pipelines inside the furnace are faulty, and can repair them in time after discovering abnormal hot spots or abnormal temperatures, avoiding production interruption and safety problems.

[0004] Therefore, real-time monitoring of the temperature inside the incinerator through infrared images is conducive to improving incineration efficiency and ensuring the safety of waste incineration.

[0005] However, due to the diffusion attenuation of infrared radiation energy and the imaging ability limit of the infrared sensor array, the resolution of infrared images is limited. The imaging results of the target usually show weak characteristics (i.e., few pixels, weak intensity, and low contrast), while the background is complex and blurred, and mixed with more clutter components. Therefore, in the prior art, through superpixel segmentation of infrared images, non-target components are filtered as much as possible, and potential target components are identified at the same time.

[0006] In related technologies, for example, a patent application document with the publication number CN116109662A discloses a superpixel segmentation method for infrared images, which discloses obtaining the position of the maximum smoothness in the scale of a specified superpixel grid by calculating image smoothness as the starting position of the superpixel clustering center, performing clustering by setting segmentation control parameters, and discriminating background pixels and target pixels, and updating the clustering result until the superpixel segmentation result is obtained.

[0007] When performing superpixel segmentation in the above solution, the number of seed points is set artificially; however, real-time images in different shooting environments may vary greatly. For example, when there are few details in the image, the difference of the whole image is small at this time, and a smaller number of seed points can meet the segmentation effect of the image. But if the set number of seed points is too large, it will increase the computational complexity of the segmentation, and the segmentation effect is not good; another example is when there are extremely rich details in the image. At this time, if the number of seed points is set small, the segmentation result will be relatively rough and cannot accurately reflect the details in the image.

[0008] Therefore, when performing superpixel segmentation of an image, the reasonable setting of the number of seed points affects the segmentation effect of the image, and further affects the problem of temperature monitoring in the incinerator. Summary of the Invention

[0009] The object of the present invention is to provide a waste feeding furnace monitoring system and method based on image processing to solve the problem in the prior art that infrared image segmentation cannot be accurately performed, resulting in inaccurate temperature monitoring in the incinerator; for this purpose, the present invention provides solutions in the following two aspects.

[0010] In the first aspect, a waste feeding furnace monitoring method based on image processing provided by the present invention includes:

[0011] Dividing the obtained infrared image of the incinerator during the incineration process to obtain a plurality of sub-regions;

[0012] Using the center of each sub-region as a seed point to perform superpixel segmentation on the infrared image to obtain a plurality of superpixel blocks; calculating the variance of the temperature means within all superpixel blocks; when the variance is greater than a threshold value, the temperature distribution in the furnace is uneven, and an alarm is given;

[0013] The sub-regions are: obtaining the closed edges and non-closed edges in the infrared image; dividing the non-closed edges into important edges and non-important edges; extending all important edges until they intersect with the boundary of the infrared image or any edge and then stop extending; the closed edges and the extended important edges divide the infrared image into a plurality of parent regions;

[0014] If the unevenness degree of the parent region is less than the set value, then this parent region is used as a sub-region; if the unevenness degree of the parent region is greater than or equal to the set value, then the remaining region except the dense region is used as a new parent region, and continue to calculate the new unevenness degree until the new unevenness degree is less than the set value and then stop, to obtain all sub-regions of each parent region; the unevenness degree is: ; , , are respectively the entropy, area, and number of edge points of the gray level of the i-th parent region, , , are respectively the entropy, area, and number of edge points of the gray level of the dense region within the i-th parent region; the dense region is the region with the largest aggregation density of non-important edges within the parent region.

[0015] The above solution first filters out the non-closed edges in the infrared image and analyzes the non-closed edges to obtain important edges. Then, through the extended important edges and closed edges, a rough segmentation of the infrared image is performed to obtain the parent regions. Next, the non-important edges within each parent region are analyzed to determine the unevenness degree of each parent region, so as to achieve the secondary fine segmentation of each parent region, and finally the child regions are obtained. That is, by dividing the obtained infrared image of the incinerator twice, the divided child regions can be accurately obtained, and taking the center of each child region as the seed point, superpixel segmentation can be performed to obtain accurate superpixel blocks, providing accurate data for the temperature monitoring of the incinerator during the incineration process.

[0016] Optionally, the specific process of extending all important edges is as follows:

[0017] Obtain the fitting curve of the edge segment corresponding to the minimum distance between any endpoint on each important edge and its corner points;

[0018] Taking the endpoint on the edge segment as the starting point, expand along the extension direction of the fitting curve.

[0019] Optionally, the dense region is the region with the largest aggregation density of non-important edges within the parent region, including:

[0020] Obtain all non-important edges within each parent region;

[0021] Adopt the mean shift clustering algorithm to cluster all non-important edges to obtain multiple class clusters;

[0022] Obtain all the outermost edge points within each class cluster, and take the region enclosed by the successively connected outermost edge points as the closed region; the outermost edge point is the edge point farthest from the cluster center on the ray where the line connecting the cluster center and any edge point is located;

[0023] Take the ratio of the number of edge points in the class cluster to the area of the corresponding closed region as the aggregation density;

[0024] Select the closed region with the largest aggregation density as the dense region of the corresponding parent region.

[0025] The above solution can accurately obtain the dense regions within each parent region.

[0026] Optionally, the process of obtaining the important edges and non-important edges is as follows:

[0027] Adopt the K-means algorithm to cluster all non-closed edges to obtain the first cluster and the second cluster; the first cluster is the cluster with a larger average length of non-closed edges, and the second cluster is the cluster with a larger average length of non-closed edges;

[0028] All non-closed edges in the first cluster and non-closed edges in the second cluster with an importance level greater than or equal to a set threshold are denoted as important edges; non-closed edges in the second cluster with an importance level less than the set threshold are denoted as unimportant edges; the importance level is positively correlated with the length and gradient of each edge in the second cluster.

[0029] The above solution can accurately screen out important edges among non-closed edges.

[0030] Optionally, the importance level is:

[0031] ; represents the importance level of the edge within the second cluster, is the length of the edge , is the minimum length of the edge within the first cluster, represents the edge at the th pixel point on the edge within the second cluster, is the total number of pixel points on the edge is the maximum value of the average gradient values of all edges.

[0032] The above solution provides a method for accurately calculating the importance level.

[0033] Optionally, it further includes that when performing superpixel segmentation, taking any pixel point as the target point, the distance from the target point to the seed point is calculated as:

[0034] ;

[0035] where , , are the distance from the target point to the seed point, the color distance, and the spatial distance respectively, is the maximum color distance within the class, is the maximum spatial distance within the class, is the importance level of the jth edge point on the line connecting the target point to the seed point, t is the total number of edge points on the line, represents the spatial distance from the jth edge point on the line to the target point;

[0036] where, when the edge where the edge point is located belongs to the first cluster, its importance level is 1; when the edge where the edge point is located belongs to the second cluster, its importance level is the importance level of the edge where the edge point is located.

[0037] In the above solution, by introducing the importance degree and the spatial difference distance from the target point to the seed point, the spatial distance from the target point to the seed point is corrected, so that the target points that cross multiple edges or edges with a greater degree of importance are not assigned to the same region as much as possible. On the contrary, they are assigned to the same region as much as possible.

[0038] Optionally, the process of obtaining the important edges and unimportant edges is as follows:

[0039] Obtain the lengths of non-closed edges, and mark the edges with a length greater than or equal to the first length threshold as important edges; mark the edges with a length less than the first length threshold as unimportant edges.

[0040] Optionally, the closed edges and non-closed edges are obtained by the Canny edge detection algorithm.

[0041] In a second aspect, a waste material feeding furnace monitoring system based on image processing includes:

[0042] A processor;

[0043] A memory storing computer instructions for waste material feeding furnace monitoring based on image processing. When the computer instructions are run by the processor, the system executes the above-mentioned waste material feeding furnace monitoring method based on image processing.

[0044] The beneficial effects of the present invention are as follows:

[0045] The solution of the present invention can make full use of the detail and edge information in the image, realize the adaptive setting of the number and position of seed points, enable the superpixel segmentation algorithm to better adapt to the edge characteristics of the image, improve the accuracy and efficiency of the superpixel segmentation algorithm, maintain edge continuity, reduce the occurrence of over-segmentation and under-segmentation phenomena, and thus provide accurate temperature data for subsequent temperature monitoring in the incinerator. Description of the Drawings

[0046] By reading the following detailed description with reference to the drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present invention will become readily understood. In the drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein:

[0047] Figure 1 Schematically shows the flowchart of the steps of a waste material feeding furnace monitoring method based on image processing in this embodiment;

[0048] Figure 2 Schematically shows the structural block diagram of a waste material feeding furnace monitoring system based on image processing in this embodiment. Detailed Embodiments

[0049] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative efforts fall within the protection scope of the present invention.

[0050] As Figure 1 shown, a waste feeding furnace monitoring method based on image processing in this embodiment includes the following steps:

[0051] Step S1, obtain the infrared image of the incinerator during the process of incinerating waste. Specifically, use an infrared camera to capture the infrared image inside the incinerator in real time.

[0052] After obtaining the infrared image, perform grayscale processing and denoising processing on the infrared image to obtain the processed infrared image.

[0053] Step S2, divide the infrared image into regions to obtain multiple sub-regions.

[0054] Among them, the process of obtaining multiple sub-regions includes steps S21 - S25, which are specifically as follows:

[0055] Step S21, obtain multiple edges in the infrared image.

[0056] In this embodiment, the Canny edge detection algorithm is used for edge detection to obtain multiple edges, and among them, the multiple edges include closed edges and non-closed edges.

[0057] Step S22, divide all non-closed edges to obtain important edges and unimportant edges.

[0058] Since the closed edges form closed regions, only non-closed edges are analyzed in this embodiment.

[0059] In one embodiment, the method for obtaining important edges and unimportant edges is:

[0060] Obtain the lengths of all non-closed edges, and mark the edges with lengths greater than or equal to the first length threshold as important edges; mark the edges with lengths less than the first length threshold as unimportant edges.

[0061] The above first length threshold is the third quartile; of course, it can also be determined according to the actual situation.

[0062] The above third quartile is obtained by obtaining the lengths of all non-closed edges, sorting them from largest to smallest, and selecting the 75% value in the sorting as the third quartile.

[0063] The length of the above non-closed edge is the total number of pixel points on the edge.

[0064] In another embodiment, the method for obtaining important edges and unimportant edges may also be:

[0065] First, cluster all non-closed edges to obtain a first cluster and a second cluster; the first cluster is the cluster with a larger average edge length, and the second cluster is the cluster with a larger average edge length.

[0066] Specifically, cluster the lengths of all non-closed edges through the K-means clustering algorithm; since the K-means clustering algorithm is a prior art, it will not be elaborated here too much.

[0067] To avoid the occurrence of a situation where the number of edges in each cluster is too sparse during clustering, the number of edges in the first cluster and the second cluster should theoretically be no less than 3.

[0068] It should be noted that the "first" and "second" above are only used to distinguish the two clusters and do not have a special defined meaning.

[0069] Secondly, calculate the importance degree of each non-closed edge in the second cluster.

[0070] Specifically, the importance degree is:

[0071] ; represents the importance degree of the edge in the second cluster, is the length of the edge , is the minimum length of the edges in the first cluster, represents the gradient value of the -th pixel point on the edge in the second cluster, is the total number of pixel points on the edge in the second cluster, is the maximum value of the average gradient values of all edges.

[0072] Since the length of any non-closed edge in the first cluster is greater than that of each non-closed edge in the second cluster, so is less than 1. When the length of the edge and the average gradient value are larger, it indicates that the isolation of the edge is weaker, and the possibility that the edge is noise is smaller. Thus, the possibility that the edge is a real texture or edge in the image is larger, and thus the importance degree of the target edge is larger.

[0073] Meanwhile, the edge length and the average gradient value The larger they are, the more it can indicate that the edge has strong continuity, and the greater the difference between the regions on both sides of the edge is, then the edge is more likely to reflect the region boundary features in the image, and the more important the target edge is. Then, all non-closed edges in the first cluster and non-closed edges in the second cluster with an importance level greater than or equal to a set threshold are marked as important edges; non-closed edges in the second cluster with an importance level greater than or equal to the set threshold are marked as unimportant edges.

[0074] In this embodiment, all edges in the first cluster are marked as important edges; at the same time, edges in the second cluster with an importance level greater than or equal to the set threshold are also marked as important edges.

[0075] The above set threshold is set to 0.7; of course, it can be set according to the actual situation.

[0076] The purpose of the above division of non-closed edges is that when the length of a non-closed edge is large (i.e., it belongs to the first cluster or the second cluster), this edge may belong to a real texture or edge, so it is relatively important; while non-closed edges with a relatively small length may be noise or other information.

[0077] Step S23: Extend each important edge until it intersects with the image boundary or any edge, and then stop extending. The different extended important edges and closed edges divide the infrared image into multiple closed regions, and each closed region is recorded as a parent region.

[0078] Specifically, the method for extending each important edge is as follows:

[0079] Obtain the fitting curve of the edge segments on each important edge, and start from the upper endpoint of the edge segment and extend along the direction of the fitting curve. The edge segment is the local edge corresponding to the minimum distance between any endpoint on each important edge and all its corner points. The corner points are obtained through the Harris corner detection algorithm.

[0080] Exemplarily, when there are two endpoints on an edge, the edge segments between the two endpoints and their nearest corner points need to be obtained respectively, that is, two endpoints correspond to two edge segments.

[0081] In this embodiment, the least squares method is used to perform curve fitting on the edge segments to obtain a fitted curve. All the obtained endpoints are iteratively extended simultaneously. Each iteration process is to include the next pixel point of the fitted curve at the endpoint into the edge, with an extension step size of 1 pixel point, and the newly extended pixel point is used as the endpoint for the next iteration; when the extended edge reaches the image boundary or intersects with any edge, the extension of the edge is stopped.

[0082] As another implementation, the condition for stopping the extension can also be that when the extended edge enters the window of any pixel point on any edge, the extension is stopped. At this time, the point on any edge that is closest to the endpoint is included in the edge, that is, the point on any edge that is closest to the endpoint is the next pixel point.

[0083] When the extension of all important edges stops, that is, when the entire extension process stops, the extended important edges and closed edges in the entire image divide the image into closed regions. Thus, the first regional division of the entire image is completed, and multiple parent regions are obtained.

[0084] Step S24, calculate the unevenness degree of each parent region.

[0085] The process of obtaining the unevenness degree in this embodiment is as follows:

[0086] First, obtain the dense regions within each parent region.

[0087] Among them, the dense regions are obtained by using the mean shift clustering algorithm to cluster all non-important edges within each parent region to obtain multiple category clusters, and the closed region of the category cluster with the largest clustering density obtained is used as the dense region.

[0088] When performing clustering, calculate the centroid of the non-important edges inside each parent region, and use the mean shift clustering algorithm for all non-important edges to perform clustering according to the position information of the centroid (the horizontal and vertical coordinate values of the centroid). Since the mean shift clustering algorithm is a prior art, it will not be elaborated here too much.

[0089] In one embodiment, the closed region of the category cluster with the largest clustering density obtained is specifically:

[0090] Obtain all the outermost edge points within each category cluster, and use the region enclosed by the sequentially connected outermost edge points as the closed region; use the ratio of the number of edge points in the category cluster to the area of the corresponding closed region as the aggregation density; select the closed region when the aggregation density is the largest as the dense region.

[0091] The above-mentioned outermost edge points are the edge points that are the farthest from the cluster center on the ray connecting the cluster center and any edge point of any edge. At this time, there are no edge points in the extending direction of the farthest edge point.

[0092] The above-mentioned enclosed area is actually the maximum circumscribed polygon of the category cluster, that is, the area surrounded by the maximum circumscribed polygon enclosing all non-important edges in the category cluster, and combined with the number of all edge points in the category cluster to obtain the aggregation density of each category cluster.

[0093] Exemplarily, when the ray where the line connecting the cluster center and the edge point a on edge A also intersects the edge point b on edge B, the distance between the edge point b and the cluster center is relatively large, and at this time, the edge point b is the outermost edge point.

[0094] The above area is the number of all pixel points within the enclosed area.

[0095] In another embodiment, the enclosed area of the category cluster with the largest obtained clustering density can also be the area enclosed by the circumscribed polygon of the category cluster when the number of edge points in each category cluster is the largest.

[0096] The above dense area needs to contain at least 30% of non-important edges, which is to avoid the dense area being too small and causing over-segmentation.

[0097] Secondly, according to the parent area and the corresponding dense area, the unevenness degree of the corresponding parent area is obtained.

[0098] In this embodiment, the unevenness degree within each parent area is calculated according to entropy and edge information density , and the formula is as follows:

[0099] ; , , are respectively the entropy, area, and number of edge points of the gray level of the i-th parent area, , , are respectively the entropy, area, and number of edge points of the gray level of the dense area within the i-th parent area.

[0100] The above entropy is the entropy of the gray level of the obtained parent area or dense area. The gray level can be dividing the gray value from 0 to 255 into 16 gray levels, that is, dividing by 16 gray values as 1 gray level. Exemplarily, gray level 1 (gray value from 0 to 15), gray level 2 (gray value from 16 to 31), and so on.

[0101] As other implementation manners, the gray level can also be the gray value, that is, the gray value of a pixel point corresponds to one gray level.

[0102] represents the edge information density of the dense area within the i-th parent area, represents the edge information density of the i-th parent area.

[0103] When is larger, it indicates that there is a significant difference in the degree of distribution disorder of the gray levels between the dense region and the parent region, suggesting that the dense region may have independent characteristics different from the corresponding parent region. At this time, the degree of non-uniformity of the parent region is larger. When is larger, it means that the dense region contains more edge information in a smaller area, and the edge information density within the dense region is greater than that of the parent region, indicating that the degree of non-uniformity of the parent region is larger.

[0104] Step S25: Based on the degree of non-uniformity of the parent region, obtain multiple sub-regions.

[0105] Specifically, the steps of obtaining multiple sub-regions include steps S251 - S253, specifically as follows:

[0106] Step S251: If the degree of non-uniformity of the parent region is less than the set value, then regard this parent region as a sub-region;

[0107] Step S252: If the degree of non-uniformity of the said parent region is greater than or equal to the set value, then regard the dense region as a sub-region, and regard the remaining region except the dense region as a new parent region; continue to calculate the degree of non-uniformity of the new parent region;

[0108] Step S253: If the new degree of non-uniformity is still greater than or equal to the set value, then repeat step S252; iterate multiple times until the final degree of non-uniformity is less than the set value and then stop to obtain all sub-regions.

[0109] Exemplarily, when , it is considered that the corresponding parent region needs to be further divided, and the current parent region is divided into two parts, namely the dense region and the remaining region.

[0110] The above set value is 0.3. Of course, it can also be set according to the actual situation.

[0111] The above calculation of the degree of non-uniformity is to divide the initial parent region multiple times until the parent region has no need for further division, that is , stop the iteration, and the divided sub-regions can be obtained. Thus, the division of sub-regions within the parent region is completed, that is, the secondary region division of the entire image is completed.

[0112] Among them, the sub-regions include the parent regions that do not need to be divided and the sub-regions within the divided parent regions.

[0113] Step S3: Using the center of each sub-region as a seed point, perform superpixel segmentation on the infrared image to obtain multiple superpixel blocks; calculate the variance of the temperatures within all the superpixel blocks; when the variance is greater than the threshold, the temperature distribution in the furnace is uneven, and an alarm is issued.

[0114] In this embodiment, using the center of each sub-region as a seed point, perform superpixel segmentation on the infrared image to obtain multiple superpixel blocks.

[0115] Specifically, the specific process of performing superpixel segmentation on the infrared image is as follows:

[0116] First, use the center of each sub-region as a seed point. Set a seed point within each sub-region, and the position of the seed point is the centroid (center) of the region. However, the color information of the seed point is determined by the average color information of all the pixel points within the sub-region excluding the edge points.

[0117] Second, for the seed point of each sub-region, use the K-means algorithm to classify the pixels around it into this region.

[0118] Among them, each pixel point that is being calculated for the distance to a certain seed point is called a target point. Modify the spatial distance according to the degree of crossing the edge of the line connecting the target point to a certain seed point. That is, the calculation formula for the distance from the target point to this seed point is as follows:

[0119] ;

[0120] Among them, , , respectively represent the distance from the target point to the seed point, the color distance, and the spatial distance. represents the maximum color distance within the class. represents the maximum spatial distance within the class. is the importance of the j-th edge point on the line connecting the target point to the seed point, t is the total number of edge points on this line. represents the spatial distance from the j-th edge point on this line to the target point, and min( ) is the minimum value function.

[0121] Among them, is the distance metric formula in the existing superpixel segmentation algorithm. Since the calculation methods and meanings of , , , are the same as those in the existing superpixel segmentation algorithm, they will not be elaborated here in detail.

[0122] Among them, when the edge where the edge point is located belongs to the first cluster, its importance is 1; when the edge where the edge point is located belongs to the second cluster, its importance is the importance degree of the edge where the edge point is located.

[0123] In the above distance metric formula, the degree of crossing edges is added. The spatial distance is corrected. When there are multiple edge pixels on the line connecting the target point to the seed point, or there is an edge point with a relatively large importance degree on the line, it indicates that there are multiple edges or an edge with a relatively large importance degree between the target point and the seed point. That is, if the cost of dividing the target point into the area of this seed point is relatively large, and it is necessary to cross multiple edges or an edge with a relatively large importance degree, it indicates that the possibility that the target point and this seed point do not belong to the same area is relatively large. At this time, the spatial distance from the target point to the seed point is increased, so that the distance from the target point to the seed point is relatively large, thereby reducing the possibility that the target point is assigned to the area of this seed point.

[0124] Then, superpixel merging is performed. Specifically, according to the distance from the pixel point to the nearest seed point and the similarity between adjacent superpixels, the superpixel merging operation is carried out.

[0125] Furthermore, iterative optimization is carried out to obtain multiple superpixel blocks. Repeat the above steps until the preset number of iterations is reached or convergence occurs, and multiple superpixel blocks after division are obtained.

[0126] Thus, through multiple iterative processes of superpixel segmentation, the segmentation of the entire image is completed.

[0127] In this embodiment, the average temperature of each region obtained after segmentation is statistically calculated, and according to the variance of the average temperatures of all regions, it is judged whether the temperature distribution in the furnace is uniform. The formula is as follows:

[0128] ; is the variance of the average temperatures of all superpixel blocks, is the th average temperature of the superpixel block, is the average temperature value of the temperatures within all superpixel blocks, is the number of superpixel blocks.

[0129] After obtaining the variance of the average temperature of the infrared image, the variance is compared with the temperature threshold to determine whether the temperature distribution in the furnace is uniform. Specifically, when the variance is greater than the temperature threshold, it is proved that the temperature distribution in the furnace is not uniform.

[0130] The value of the above temperature threshold can be 300°, and of course, it can also be set according to the actual situation.

[0131] In this embodiment, when the temperature distribution in the furnace is uneven, an alarm is also required to promptly remind the staff that abnormal conditions may occur in the furnace for maintenance.

[0132] The solution of the present invention divides the infrared image twice to obtain multiple regions, and then corrects the spatial distance by using the degree of crossing the edge of the connection line between the target point and the seed point in each region, calculates the distance from the target point to the seed point, and performs multiple iterations to complete the entire superpixel segmentation. According to the variance of the temperatures of all regions, it is judged whether the temperature distribution in the furnace is uniform.

[0133] The present invention also provides a waste material feeding into furnace monitoring system based on image processing. As Figure 2 shown, the monitoring system includes a processor and a memory, and the memory stores computer program instructions, which when executed by the processor implement a waste material feeding into furnace monitoring method based on image processing as described above according to the present invention.

[0134] The monitoring system further includes a communication bus and a communication interface and other components well known to those skilled in the art, and their settings and functions are known in the art, so they will not be described in detail here.

[0135] In the present invention, the aforementioned memory can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component. For example, a computer-readable storage medium can be any suitable magnetic storage medium or magneto-optical storage medium, such as resistive random access memory RRAM (Resistive Random Access Memory), dynamic random access memory DRAM (Dynamic Random Access Memory), static random access memory SRAM (Static Random-Access Memory), enhanced dynamic random access memory EDRAM (Enhanced Dynamic Random Access Memory), high-bandwidth memory HBM (High-Bandwidth Memory), hybrid memory cube HMC (Hybrid Memory Cube), etc., or any other medium that can be used to store the required information and can be accessed by an application program, module, or both. Any such computer storage medium can be a part of the device or accessible or connectable to the device. Any application or module described in the present invention can be implemented by computer-readable / executable instructions stored or otherwise held by such a computer-readable medium.

[0136] In the description of this specification, the meaning of "a plurality of" is at least two, such as two, three, or more, etc., unless otherwise specifically and clearly defined.

[0137] Although this specification has shown and described multiple embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will think of many changes, alterations, and alternative ways without departing from the spirit and scope of the present invention. It should be understood that various alternatives to the embodiments of the present invention described herein may be employed in practicing the present invention.

Claims

1. A method for monitoring waste entering a furnace based on image processing, characterized in that: include: Divide the acquired infrared image of the incinerator during the incineration process to obtain multiple sub-areas; Taking the center of each sub-region as a seed point, the infrared image is segmented into superpixels to obtain multiple superpixel blocks; the variance of the temperature mean in all superpixel blocks is calculated; when the variance is greater than a threshold, the temperature distribution in the furnace is uneven, and an alarm is issued; The sub-regions are: obtaining closed edges and non-closed edges in the infrared image; dividing the non-closed edges into important edges and non-important edges; extending all important edges until they intersect with the boundary or any edge of the infrared image and stop extending; the closed edges and the extended important edges divide the infrared image into multiple parent regions; If the unevenness of the parent area is less than the set value, the parent area is used as the child area; if the unevenness of the parent area is greater than or equal to the set value, the remaining area except the dense area is used as the new parent area, and the new unevenness calculation is continued until the new unevenness is less than the set value, and all the child areas of each parent area are obtained; the unevenness for: ; , , are the grayscale entropy, area, and number of edge points of the i-th parent region, respectively. , , are respectively the grayscale entropy, area, and number of edge points of the dense area in the i-th parent area; the dense area is the area with the largest density of non-important edges in the parent area.

2. The method for monitoring waste entering a furnace based on image processing according to claim 1, characterized in that: The specific process of extending all important edges is: Obtain the fitting curve of the edge segment corresponding to the minimum distance between any endpoint on each important edge and each corner point on it; Taking the endpoint on the edge segment as the starting point, expand along the extension direction of the fitting curve.

3. The method for monitoring waste entering a furnace based on image processing according to claim 1, characterized in that: The dense area is the area with the largest density of non-important edges in the parent area, including: Get all non-significant edges in each parent region; The mean shift clustering algorithm is used to cluster all non-important edges and obtain multiple category clusters; Obtain all outermost edge points in each category cluster, and use the area enclosed by the outermost edge points connected in sequence as a closed area; the outermost edge point is the edge point farthest from the cluster center on the ray where the cluster center and any edge point are connected; The ratio of the number of edge points in the category cluster to the area of ​​the corresponding closed region is taken as the clustering density; The closed area with the largest aggregation density is selected as the dense area corresponding to the parent area.

4. The method for monitoring waste entering a furnace based on image processing according to claim 1, characterized in that: The process of obtaining the important edges and the unimportant edges is as follows: The K-means algorithm is used to cluster all non-closed edges to obtain the first cluster and the second cluster; the first cluster is the cluster with a larger mean length of non-closed edges, and the second cluster is the cluster with a larger mean length of non-closed edges; All non-closed edges in the first cluster and non-closed edges in the second cluster whose importance is greater than or equal to a set threshold are recorded as important edges; The non-closed edges in the second cluster whose importance is less than a set threshold are recorded as non-important edges; the importance is positively correlated with the length and gradient of each edge in the second cluster.

5. The method for monitoring waste entering a furnace based on image processing according to claim 4, characterized in that: The importance levels are: ; represents the inner edge of the second cluster The importance of For the edge Length, is the minimum length of the inner edge of the first cluster, represents the inner edge of the second cluster Previous The gradient value of a pixel, The inner edge of the second cluster The total number of pixels on is the maximum value of the average gradient of all edges.

6. The method for monitoring waste entering a furnace based on image processing according to claim 5, characterized in that: It also includes that when performing superpixel segmentation, any pixel point is taken as the target point, and the distance from the target point to the seed point is calculated as: ； in, , , They are the distance, color distance, and spatial distance from the target point to the seed point, respectively. is the maximum color distance within the class, is the maximum spatial distance within the class, is the importance of the jth edge point on the line connecting the target point to the seed point, t is the total number of edge points on the line, represents the spatial distance from the jth edge point on the line to the target point; When the edge where the edge point is located belongs to the first cluster, its importance is 1; when the edge where the edge point is located belongs to the second cluster, its importance is the importance of the edge where the edge point is located.

7. The method for monitoring waste entering a furnace based on image processing according to claim 1, characterized in that: The process of obtaining the important edges and the unimportant edges is as follows: The length of each non-closed edge is obtained, and the edge whose length is greater than or equal to the first length threshold is recorded as an important edge; the edge whose length is less than the first length threshold is recorded as a non-important edge.

8. The method for monitoring waste entering a furnace based on image processing according to claim 1, characterized in that: The closed edges and the non-closed edges are obtained by using the Canny edge detection algorithm.

9. A waste furnace monitoring system based on image processing, characterized in that: include: processor; A memory storing computer instructions for monitoring waste entering a furnace based on image processing. When the computer instructions are executed by the processor, the system executes a method for monitoring waste entering a furnace based on image processing according to any one of claims 1 to 8.

Images