A bar-by-bar tracking method for sawing based on multi-target video tracking

Abstract

The present invention discloses a method for tracking each bar during sawing based on multi-target video tracking, including: establishing the underlying logic for tracking the bars based on roller signals, cold detection signals and sawing signals; establishing a single bar tracking model based on a multi-target video tracking system, each bar including three attributes of position, length and number, wherein the bar number is its unique ID; when the bar is sawn by a grinding wheel saw, the bar number is automatically updated according to a given numbering rule; and judging and alarming the bars that have logical anomalies during the tracking process. The method of the present invention realizes the tracking of the bars in the sawing area one by one by specifying the bars and their unique numbers, which effectively solves the problem of confusing logic in tracking the bars in the sawing area. The method of the present invention has clear logic, high accuracy and reliability, and is easy to maintain without changing the original production line.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent manufacturing of steel products, and in particular to a bar sawing and tracking method based on multi-target video tracking. Background Art

[0002] In recent years, the production concentration of the domestic bar industry has increased significantly, and it has gradually developed in the direction of scale and specialization. At the same time, it also faces many problems. Among them, it is particularly important to establish a complete quality traceability system for the production process. At present, many domestic companies' batch tracking and control methods can no longer meet the actual needs of refined production. When problems occur in the performance of finished bar products, bar manufacturers hope to trace back to the specific bar and the specific process where the problem occurs, so as to analyze the problem and make continuous improvements.

[0003] In the process of bar production, unqualified products may be thrown off the line and work may be reversed. These factors make the logical order of the bar blanks easily confused, so it is difficult to achieve manual participation and PLC-based bar tracking one by one. In addition, the existing bar-by-bar tracking methods are mostly based on marking and identification, which are not suitable for tracking the bar sawing area on the production line.

[0004] Sawing is one of the processes with the most complex tracking logic on the bar production line. The bars move in batches in the sawing area, and multiple bars may be sawed at the same time, side by side, or continuously. The existing bar tracking methods are mostly based on identification means, and it is impossible to determine the status of each bar and the source of each small section. Once quality problems occur, it is difficult to trace. Summary of the invention

[0005] The purpose of the present invention is to provide a method for tracking each bar during sawing based on multi-target video tracking, so as to solve the problem that the existing method for tracking each bar during sawing based on identification means is difficult to achieve tracking of multiple bars being sawed side by side at the same time.

[0006] A bar sawing tracking method based on multi-target video tracking comprises the following steps:

[0007] S1: Establish the underlying logic for tracking the bars based on roller signals, cold inspection signals and sawing signals;

[0008] S2: Based on the multi-target video tracking system, a single bar tracking model is established. Each bar contains three attributes: position, length, and number, where the bar number is its unique ID;

[0009] S3: When the bar is cut by the grinding wheel saw, the bar number is automatically updated according to the given numbering rule;

[0010] S4: Determine and alarm the bars that have logical anomalies during the tracking process.

[0011] Preferably, in step S1, the underlying logic of establishing tracking of the bars based on the roller signal, the cold detection signal and the sawing signal means that when the bars pass through the sawing area, the information of the bars by batch is obtained through the roller signal, the cold detection signal and the sawing signal, including:

[0012] S11: The time and order of the bars arriving at the entrance in batches are obtained through the cold inspection signal at the entrance of the sawing area, and the batch to which the bars belong is determined by comparing with the upstream production plan, so as to obtain the initial number of the bars;

[0013] S12: The roller signal is used to determine whether the bar is moving;

[0014] S13: Obtain sawing time and sawing position through sawing signals to clarify actual sawing information of each process of the bar;

[0015] S14: The time when the bars arrive at the exit is obtained through the cold detection signal at the exit of the sawing area, and the order of each sequence after the sawing of the batch of bars is formed, and the order is matched with the information of the multi-target video tracking system.

[0016] Preferably, the order of each sequence formed after the batch of bars is sawed is corresponded to the information of the multi-target video tracking system, which means that a batch of bars will generate several bar sequences after sawing. The multi-target video tracking system is interlocked with the underlying logic signal to clearly identify the sequence of bars arriving at the exit of the sawing area, and outputs the bar numbers in the sequence and information on abnormal conditions.

[0017] Preferably, in step S2, the working method of the multi-target video tracking system is as follows:

[0018] S21: Multiple industrial cameras are set up on the production line of the sawing area. The viewfinders of adjacent industrial cameras overlap. A synchronizer is used to synchronize the multiple industrial cameras to achieve full coverage of the bar production line.

[0019] S22: stitching the obtained synchronized images to obtain a global image of the sawing area;

[0020] S23: performing target detection on the obtained global image to obtain a detection result of each bar instance;

[0021] S24: Target matching and number matching are performed on the detected rod material instance in the current frame and the rod material instance in the previous frame.

[0022] Preferably, the single bar tracking model is as shown in formula (1):

[0023] Bar((x,y),l,C) (1)

[0024] Among them, (x, y) represents the position coordinates of the center of the two-dimensional geometric model of a single bar in the video screen, which corresponds to the actual position on the roller, and the positive direction of the x-axis is the moving direction of the bar; l represents the length of the bar; C represents the bar number.

[0025] Preferably, in step S3, when the bar is sawn by the grinding wheel saw, the bar number is automatically updated according to a given numbering rule, which means that when the bar is sawn by the 1# grinding wheel saw, the 2# grinding wheel saw, ..., the N# grinding wheel saw in the sawing area, a new bar will be logically generated, and the sawed bar needs to be automatically numbered, and the tracking of each bar is achieved by specifying the bar and its unique number, as shown in formula (2):

[0026] C1=C0+Na,C2=C0+Nb (2)

[0027] Among them, C0 represents the original number of the rod; C1 and C2 represent the new numbers of the two rods generated after the rod numbered C0 is divided into two; N represents that the rod is cut by an N# grinding wheel saw; a and b represent the parameters used to distinguish the two rods.

[0028] Preferably, in step S4, the logic abnormality includes:

[0029] Stuck, that is, the bar is stuck on the roller, stops moving, and disrupts normal production;

[0030] Lost bar, that is, the bar rolls directly out of the roller table, causing the bar to come off the line abnormally;

[0031] Falling behind means that the rod moves slowly due to reasons including friction, and falls behind from the previous sequence to the next sequence, causing confusion in the tracking logic.

[0032] Preferably, in step S4, the determination and alarm of the bar material having a logical abnormality during the tracking process includes:

[0033] S41: If at a certain moment the roller signal determines that the bar is moving, and the position of a certain bar does not change, as shown in formula (3):

[0034] Δx=0 (3)

[0035] Where Δx represents the lateral displacement of the rod, it is judged as a stuck rod. At this time, the multi-target video tracking system issues a stuck rod alarm, outputs the information of the stuck rod, and performs manual intervention operations;

[0036] S42: If the position of a certain bar reaches the two marking lines set on both sides of the global screen to mark the roller range, that is, the edge control lines, as shown in formula (4):

[0037] y≤y T1 or y≥yT2 (4)

[0038] Where y is the ordinate of the rod support position, T1 With y T2 If the vertical coordinates of the two edge control lines are the same, it is judged as a lost bar. At this time, the multi-target video tracking system issues a lost bar alarm, outputs the information of the lost bar, and performs manual intervention operations;

[0039] S43: If the geometric center position of a bar lags behind the geometric center positions of other bars in its sequence by at least l, as shown in formula (5):

[0040] x n -x>l (5)

[0041] Where x is the horizontal coordinate of the geometric center position of the rod, n is the horizontal coordinate of the geometric center position of other bars in the same sequence, and l is the length of the bar in the sequence. It is judged as falling behind. At this time, the multi-target video tracking system issues a falling-behind alarm, outputs the falling-behind bar information, and performs manual intervention operations.

[0042] The beneficial effects brought about by the technical solution provided by the embodiment of the present invention include at least:

[0043] The method for tracking each bar during the sawing process based on multi-target video tracking provided by the present invention realizes tracking each bar by specifying the bar and its unique number. The bars move in batches in the sawing area. Most of the existing methods for tracking each bar are based on identification means, which cannot determine the status of each bar and the source of each small section, and it is difficult to trace the quality data in the production process. The method of the present invention realizes automatic numbering and tracking of bars in the sawing area, and effectively solves the problem of confusing logic in tracking bars in the sawing area. The method of the present invention has clear logic, high accuracy and reliability, and is easy to maintain without changing the original production line. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0045] Figure 1 is a flow chart of a bar-by-bar tracking method for sawing based on multi-target video tracking provided by an embodiment of the present invention;

[0046] Figure 2a-2d It is a schematic diagram of the time T, time T+1, time T+2, and time T+3 in the process of numbering newly added bars provided by an embodiment of the present invention;

[0047] Figure 3 is a detailed flowchart of a multi-target video tracking method provided by an embodiment of the present invention;

[0048] Figure 4 is a schematic diagram of the installation position of the synchronous camera provided in an embodiment of the present invention;

[0049] Figure 5a-Figure 5b It is a schematic diagram of time T and time T+1 in the case of a logical abnormality in the bar tracking process provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] The embodiment of the present invention provides a bar sawing tracking method based on multi-target video tracking, such as Figure 1 As shown, the method comprises the following steps:

[0052] S1: The underlying logic for tracking the bars is established based on roller signals, cold inspection signals and sawing signals.

[0053] In this step, the underlying logic of establishing bar tracking based on roller signals, cold inspection signals and sawing signals refers to obtaining bar batch information through roller signals, cold inspection signals and sawing signals when the bars pass through the sawing area, including:

[0054] S11: The time and order of the bars arriving at the entrance in batches are obtained through the cold inspection signal at the entrance of the sawing area, and the batch to which the bars belong is determined by comparing with the upstream production plan, so as to obtain the initial number of the bars;

[0055] S12: The roller signal is used to determine whether the bar is moving;

[0056] S13: Obtain sawing time and sawing position through sawing signals to clarify actual sawing information of each process of the bar;

[0057] S14: The time when the bars arrive at the exit is obtained through the cold detection signal at the exit of the sawing area, the order of each sequence after the sawing of the batch of bars is formed, and the order is matched with the information of the multi-target video tracking system.

[0058] Among them, the order of each sequence after the batch of bars is sawed is corresponded with the information of the multi-target video tracking system, which means that a batch of bars will generate several bar sequences after sawing. The multi-target video tracking system is interlocked with the underlying logic signal to clarify the sequence of bars arriving at the exit of the sawing area, and outputs the bar number in the sequence and information on abnormal conditions.

[0059] S2: Based on the multi-target video tracking system, a single bar tracking model is established. Each bar contains three attributes: position, length, and number, among which the bar number is its unique ID.

[0060] In this step, the working process of the multi-target video tracking system is as follows:

[0061] S21: Multiple industrial cameras are set up on the production line of the sawing area. The viewfinders of adjacent industrial cameras overlap. A synchronizer is used to synchronize the multiple industrial cameras to achieve full coverage of the bar production line.

[0062] S22: stitching the obtained synchronized images to obtain a global image of the sawing area;

[0063] S23: performing target detection on the obtained global image to obtain a detection result of each bar instance;

[0064] S24: Target matching and number matching are performed on the detected rod material instance in the current frame and the rod material instance in the previous frame.

[0065] The single bar tracking model is as shown in formula (1):

[0066] Bar((x,y),l,C) (1)

[0067] Among them, (x, y) represents the position coordinates of the center of the two-dimensional geometric model of a single bar in the video screen, which corresponds to the actual position on the roller, and the positive direction of the x-axis is the moving direction of the bar; l represents the length of the bar; C represents the bar number.

[0068] S3: When the bar is cut by the grinding wheel saw, the bar number is automatically updated according to the given numbering rule.

[0069] In this step, when the bar is sawn by the grinding wheel saw, the bar number is automatically updated according to the given numbering rule, which means that when the bar is sawn by the 1# grinding wheel saw, the 2# grinding wheel saw, ..., the N# grinding wheel saw in the sawing area, a new bar will be generated logically, and the sawed bar needs to be automatically numbered, and the bar and its unique number are designated to achieve tracking one by one, as shown in formula (2):

[0070] C1=C0+Na,C2=C0+Nb (2)

[0071] Among them, C0 represents the original number of the rod; C1 and C2 represent the new numbers of the two rods generated after the rod numbered C0 is divided into two; N represents that the rod is cut by an N# grinding wheel saw; a and b represent the parameters used to distinguish the two rods.

[0072] S4: Determine and alarm the bars that have logical anomalies during the tracking process.

[0073] Specifically, the logic abnormalities include:

[0074] Stuck, that is, the bar is stuck on the roller, stops moving, and disrupts normal production;

[0075] Lost bar, that is, the bar rolls directly out of the roller table, causing the bar to come off the line abnormally;

[0076] Falling behind means that the rod moves slowly due to friction and other reasons, and falls behind from the previous sequence to the next sequence, causing confusion in the tracking logic.

[0077] In this step, the determination and alarm of the bar material having a logical abnormality during the tracking process includes:

[0078] S41: If at a certain moment the roller signal determines that the bar is moving, and the position of a certain bar does not change, as shown in formula (3):

[0079] Δx=0 (3)

[0080] Where Δx represents the lateral displacement of the rod, it is judged as a stuck rod. At this time, the multi-target video tracking system issues a stuck rod alarm, outputs the information of the stuck rod, and performs manual intervention and other operations;

[0081] S42: If the position of a certain bar reaches the two marking lines set on both sides of the global screen to mark the roller range, that is, the edge control lines, as shown in formula (4):

[0082] y≤y T1 or y≥y T2 (4)

[0083] Where y is the ordinate of the rod support position, T1 With y T2 If the vertical coordinates of the two edge control lines are the same, it is judged as a lost bar. At this time, the multi-target video tracking system issues a lost bar alarm, outputs the information of the lost bar, and performs manual intervention and other operations;

[0084] S43: If the geometric center position of a bar lags behind the geometric center positions of other bars in its sequence by at least l, as shown in formula (5):

[0085] xn -x>l (5)

[0086] Where x is the horizontal coordinate of the geometric center position of the rod, n is the horizontal coordinate of the geometric center position of other bars in the same sequence, and l is the length of the bar in the sequence. It is judged as falling behind. At this time, the multi-target video tracking system issues a falling-behind alarm, outputs the falling-behind bar information, and performs manual intervention and other operations.

[0087] Specifically, take the case where three bars are sawn side by side by three grinding wheel saws at the same time as an example. Figure 2a-2d The schematic diagram is a process diagram of numbering newly added bars. The bar sawing tracking method based on multi-target video tracking includes the following steps:

[0088] The underlying logic for tracking bars is established based on roller signals, cold inspection signals, and sawing signals, including:

[0089] Through the cold inspection signal at the entrance of the sawing area, the time and order of the bars arriving at the entrance in batches are obtained, and the bar batches are determined by comparing with the upstream production plan, and the initial bar numbers are obtained as 1, 2, and 3;

[0090] The roller signal can be used to determine whether the bar is moving;

[0091] The sawing time and sawing position are obtained through the sawing signal. The batch of bars is sawn by the 1# grinding wheel saw at time T+1, and the two bar sequences generated are sawn by the 2# grinding wheel saw and the 3# grinding wheel saw respectively at time T+3;

[0092] Through the cold detection signal at the exit of the sawing area, the time when the bars arrive at the exit is obtained, forming the order of the four bar sequences after the sawing is completed. The multi-target video tracking system is interlocked with the underlying logic signal to clarify the sequence of bars arriving at the exit of the sawing tracking area, and output the bar number in the sequence and information on abnormal conditions.

[0093] Figure 3 It is a detailed flowchart of the multi-target video tracking method. Figure 4 The schematic diagram of the synchronous camera installation position. The working method of the multi-target video tracking system is as follows:

[0094] Multiple industrial cameras are installed on the production line of the sawing tracking area. There is a certain overlap between the views of adjacent industrial cameras. A synchronizer is used to synchronize multiple industrial cameras to achieve full coverage of this section of the bar production line.

[0095] The obtained synchronized images are stitched together to obtain a global image of the sawing area;

[0096] After the bar enters the sawing area, the initial number is obtained by the production plan, and each bar instance detected in the initial frame image is numbered accordingly. In the subsequent video frame analysis, target detection is performed first, and then the current frame image detection result is matched with the previous frame result. The matched bar instances are numbered accordingly, the numbers of the sawed bar instances are updated, and the numbers of the abnormal bars are output.

[0097] The single bar tracking model established based on the multi-target video tracking system is shown in formula (1):

[0098] Bar((x,y),l,C) (1)

[0099] Among them, (x, y) represents the position coordinates of the center of the two-dimensional geometric model of a single bar in the video screen, which corresponds to the actual position on the roller, and the positive direction of the x-axis is the moving direction of the bar; l represents the length of the bar; C represents the bar number.

[0100] When the bar is sawed by 1# grinding wheel saw, 2# grinding wheel saw...N# grinding wheel saw in the sawing area, new bars will be generated logically. The sawed bars need to be automatically numbered and tracked one by one by specifying the bars and their unique numbers.

[0101] The embodiment of the present invention takes the case where three bars are sawed side by side by three grinding wheel saws as an example, and provides a rule for numbering the newly added bars, as shown in formula (2):

[0102] C1=C0+Na,C2=C0+Nb (2)

[0103] Among them, C0 represents the original number of the rod; C1 and C2 represent the new numbers of the two rods generated after the rod numbered C0 is divided into two; N represents that the rod is cut by an N# grinding wheel saw; a and b represent the parameters used to distinguish the two rods.

[0104] The specific steps are as follows:

[0105] At time T, before the 1# grinding wheel saw, the bars are initially numbered 1, 2, and 3;

[0106] At time T+1, the sawing signal of the 1# grinding wheel saw is triggered, and the newly added bars produced by the sawing of the 1# grinding wheel saw are numbered as 1-1a and 1-1b, 2-1a and 2-1b, 3-1a and 3-1b;

[0107] At time T+2, the three bars in the latter sequence arrive at the 2# grinding wheel saw, and the three bars in the former sequence directly arrive at the 3# grinding wheel saw without being sawed by the 2# grinding wheel saw;

[0108] At time T+3, the sawing signal of the 2# grinding wheel saw is triggered, and the three bars in the next sequence are sawn by the 2# grinding wheel saw. The generated new bars are numbered as 1-1a-2a and 1-1a-2b, 2-1a-2a and 2-1a-2b, and 3-1a-2a and 3-1a-2b;

[0109] At time T+3, the sawing signal of the 3# grinding wheel saw is triggered, and the three bars in the previous sequence are sawn by the 3# grinding wheel saw. The new bars are numbered as 1-1b-3a and 1-1b-3b, 2-1b-3a and 2-1b-3b, 3-1b-3a and 3-1b-3b.

[0110] Figure 5a-Figure 5b This is a schematic diagram of the logic anomalies that occur during the bar tracking process. Common logic anomalies include: sticking, that is, the bar is stuck on the roller, stops moving, and disrupts normal production; losing, that is, the bar directly rolls out of the roller, causing the bar to abnormally go offline; falling behind, that is, the bar moves slowly due to friction and other reasons, and falls behind from the previous sequence to the next sequence, causing confusion in tracking logic.

[0111] If at a certain moment the roller signal is used to determine that the bar is moving, and the position of a certain bar does not change, as shown in formula (3):

[0112] Δx=0 (3)

[0113] Where Δx represents the lateral displacement of the rod, which is judged as a stuck rod. At this time, the multi-target video tracking system issues a stuck rod alarm, outputs the information of the stuck rod, and performs manual intervention and other operations.

[0114] If the position of a certain bar reaches the two marking lines set on both sides of the global screen to mark the roller range, that is, the edge control lines, as shown in formula (4):

[0115] y≤y T1 or y≥y T2 (4)

[0116] Where y is the ordinate of the rod support position, T1 With y T2 If the vertical coordinates of the two edge control lines are the same, it is judged as a lost bar. At this time, the multi-target video tracking system issues a lost bar alarm, outputs the information of the lost bar, and performs manual intervention and other operations.

[0117] If the geometric center position of a bar lags behind the geometric center position of other bars in its sequence by at least l, as shown in formula (5):

[0118] x n -x>l (5)

[0119] Where x is the horizontal coordinate of the geometric center position of the rod, n is the horizontal coordinate of the geometric center position of other bars in the same sequence, and l is the length of the bar in the sequence. It is judged as falling behind. At this time, the multi-target video tracking system issues a falling-behind alarm, outputs the falling-behind bar information, and performs manual intervention and other operations.

[0120] In summary, the method for tracking each bar during the sawing process based on multi-target video tracking provided by the present invention realizes tracking each bar by specifying the bar and its unique number. The bars move in batches in the sawing area. Most of the existing methods for tracking each bar are based on identification means, which cannot determine the status of each bar and the source of each small section, and it is difficult to trace the quality data in the production process. The method of the present invention realizes automatic numbering and tracking of bars in the sawing area, and effectively solves the problem of confusing logic in tracking bars in the sawing area. The method of the present invention has clear logic, high accuracy and reliability, and is easy to maintain without changing the original production line.

[0121] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A bar sawing tracking method based on multi-target video tracking, characterized in that: The following steps are involved: S1: Establish the underlying logic for tracking the bars based on roller signals, cold inspection signals and sawing signals; In step S1, the underlying logic of establishing tracking of the bars based on the roller signal, the cold detection signal and the sawing signal means that when the bars pass through the sawing area, the information of the bars by batch is obtained through the roller signal, the cold detection signal and the sawing signal, including: S11: The time and order of the bars arriving at the entrance in batches are obtained through the cold inspection signal at the entrance of the sawing area, and the batch to which the bars belong is determined by comparing with the upstream production plan, so as to obtain the initial number of the bars; S12: The roller signal is used to determine whether the bar is moving; S13: Obtain sawing time and sawing position through sawing signals to clarify actual sawing information of each process of the bar; S14: obtaining the time when the bars arrive at the exit through the cold detection signal at the exit of the sawing area, forming the order of each sequence after the sawing of the batch of bars is completed, and corresponding to the information of the multi-target video tracking system; S2: Based on the multi-target video tracking system, a single bar tracking model is established. Each bar contains three attributes: position, length, and number, where the bar number is its unique ID; In step S2, the working method of the multi-target video tracking system is as follows: S21: Multiple industrial cameras are set up on the production line of the sawing area. The viewfinders of adjacent industrial cameras overlap. A synchronizer is used to synchronize the multiple industrial cameras to achieve full coverage of the bar production line. S22: stitching the obtained synchronized images to obtain a global image of the sawing area; S23: performing target detection on the obtained global image to obtain a detection result of each bar instance; S24: matching the detected rod material instance in the current frame with the rod material instance in the previous frame and making the numbers correspond; The single bar tracking model is shown in formula (1): Bar((x,y),l,C) (1) Among them, (x, y) represents the position coordinates of the center of the two-dimensional geometric model of a single bar in the video screen, which corresponds to the actual position on the roller, and the positive direction of the x-axis is the moving direction of the bar; l represents the length of the bar; C represents the bar number; S3: When the bar is cut by the grinding wheel saw, the bar number is automatically updated according to the given numbering rule; In step S3, when the bar is sawn by the grinding wheel saw, the bar number is automatically updated according to the given numbering rule, which means that when the bar is sawn by the 1# grinding wheel saw, the 2# grinding wheel saw, ..., the N# grinding wheel saw in the sawing area, a new bar will be logically generated, and the sawed bar needs to be automatically numbered, and the tracking of each bar is achieved by specifying the bar and its unique number, as shown in formula (2): C1=C0+Na,C2=C0+Nb (2) Among them, C0 represents the original number of the bar; C1 and C2 represent the new numbers of the two bars generated after the bar numbered C0 is divided into two; N represents that the bar is cut by the N# grinding wheel saw; a and b represent the parameters used to distinguish the two bars; S4: Determine and alarm the bars that have logical anomalies during the tracking process.

2. The method for tracking bar sawing one by one based on multi-target video tracking according to claim 1 is characterized in that: The order of each sequence after the batch of bars is sawed is formed, and corresponds to the information of the multi-target video tracking system, which means that after a batch of bars is sawn, several bar sequences will be generated. The multi-target video tracking system is interlocked with the underlying logic signal to clearly identify the sequence of bars arriving at the exit of the sawing area, and outputs the bar number in the sequence and information on abnormal conditions.

3. The method for tracking bar sawing based on multi-target video tracking according to claim 1 is characterized in that: In step S4, the logic abnormality includes: Stuck, that is, the bar is stuck on the roller, stops moving, and disrupts normal production; Lost bar, that is, the bar rolls directly out of the roller table, causing the bar to come off the line abnormally; Falling behind means that the rod moves slowly due to reasons including friction, and falls behind from the previous sequence to the next sequence, causing confusion in the tracking logic.

4. The method for tracking bar sawing based on multi-target video tracking according to claim 3 is characterized in that: In step S4, the determination and alarm of the bar material having a logical abnormality during the tracking process includes: S41: If at a certain moment the roller signal determines that the bar is moving, and the position of a certain bar does not change, as shown in formula (3): Δx=0 (3) Where Δx represents the lateral displacement of the rod, it is judged as a stuck rod. At this time, the multi-target video tracking system issues a stuck rod alarm, outputs the information of the stuck rod, and performs manual intervention operations; S42: If the position of a certain bar reaches the two marking lines set on both sides of the global screen to mark the roller range, that is, the edge control lines, as shown in formula (4): y≤y T1 or y≥y T2 (4) Where y is the ordinate of the rod support position, T1 With y T2 If the vertical coordinates of the two edge control lines are the same, it is judged as a lost bar. At this time, the multi-target video tracking system issues a lost bar alarm, outputs the information of the lost bar, and performs manual intervention operations; S43: If the geometric center position of a bar lags behind the geometric center positions of other bars in its sequence by at least l, as shown in formula (5): x n -x>l (5) Where x is the horizontal coordinate of the geometric center position of the rod, n is the horizontal coordinate of the geometric center position of other bars in the same sequence, and l is the length of the bar in the sequence. It is judged as falling behind. At this time, the multi-target video tracking system issues a falling-behind alarm, outputs the falling-behind bar information, and performs manual intervention operations.

Images