Defective metal can guiding and indicating system based on visual identification

Abstract

A defective metal can guiding and indicating system based on visual identification comprises a processing terminal and visual identification cameras arranged on the two sides of a conveying belt, the number of the visual identification cameras is not less than two, the visual identification cameras are higher than the conveying belt and are arranged in a downward inclined mode, and the visual identification cameras are arranged towards the conveying belt. A rectangular visual identification area is arranged at the same position of the conveying belt; a plurality of follow spot lamps are arranged below each visual identification camera, and the follow spot lamps can track target metal cans based on the target metal cans identified by the visual identification cameras under the control of the processing terminal. And the projection light of the follow spotlight synchronously moves along with the target metal can and is turned off until the target metal can exceeds the visual identification area. Through the assistance of the follow spot lamp, a worker can quickly find and identify the deformed metal can, and then the worker can identify and manually reject the metal can according to the information of the follow spot lamp.

Description

Technical Field

[0001] The invention relates to the field of metal can stacking production lines, and specifically to a defective metal can guidance and indication system based on visual recognition. Background Art

[0002] In the process of stacking and packaging metal cans, deformed metal cans are inevitable. The presence of deformed metal cans will affect the arrangement and normal transmission of other normal metal cans. The existing solution is that at the conveyor belt position, the staff can identify the deformed metal cans with the human eye and remove them from the conveyor belt. The above method can prevent the metal cans with more serious deformation from getting stuck in front of the sorting and screening structure and affecting the normal arrangement of the metal cans. As for the metal cans with less serious deformation, as the conveyor belt moves, the staff may not be able to detect them in time, which will cause them to fall normally into the subsequent packaging line, which will have a certain impact on the yield rate of the products leaving the factory. Summary of the Invention

[0003] In order to solve the above-mentioned problem that the human eye cannot easily identify deformed metal cans, the invention provides a defective metal can guidance and indication system based on visual recognition.

[0004] The technical solution of the invention is as follows:

[0005] A defective metal can guidance and indication system based on visual recognition includes a processing terminal and visual recognition cameras arranged on both sides of a conveyor belt. The number of visual recognition cameras is not less than two. The visual recognition cameras are arranged above the conveyor belt and tilted downward. The visual recognition cameras are all arranged facing the conveyor belt, and a rectangular visual recognition area is set at the same position on the conveyor belt.

[0006] A plurality of follow-up lights are provided below each visual recognition camera. Under the control of the processing terminal, the follow-up lights can track the target metal can based on the target metal can identified by the visual recognition camera, and the projection light of the follow-up lights moves synchronously with the target metal can until the target metal can exceeds the visual recognition area and then turns off;

[0007] The length of the visual recognition area in the width direction of the conveyor belt shall not be less than the width of the conveyor belt.

[0008] With the help of follow-spot lights, workers can quickly find and identify deformed metal cans. They can then identify and manually remove the metal cans based on the information from the follow-spot lights.

[0009] In order to facilitate human eye observation, the projection light color of the chasing light is green.

[0010] As a preferred solution, two visual recognition cameras are provided and symmetrically arranged on both sides of the conveyor belt and supported by a lifting bracket fixed to the conveyor belt.

[0011] In order to project light onto multiple deformed metal cans at the same time, the lifting bracket is provided with multiple follow-up lights spaced apart in the height direction below the visual recognition camera, and the follow-up lights can rotate in the horizontal plane and the vertical plane.

[0012] To prevent multiple spotlights from projecting onto the same can, all spotlights are connected to angle encoders on their horizontal axes. These encoders are connected to a processing terminal. If at least two spotlights are detected tracking the same can, the processing terminal can shut down at least one spot, retaining one spot, and removing the can's identification information from the visual recognition camera corresponding to the shut-down spot. Only a single spotlight is needed to provide guidance to personnel; using multiple spotlights simultaneously provides no additional benefit, and increasing the number of spotlights used can affect subsequent can identification.

[0013] The method for achieving the above identification is that the identification of at least two tracking lights tracking the same target metal can is specifically achieved through the following operations:

[0014] Establish a coordinate system: Establish an XY coordinate system in the visual recognition area, with the X axis parallel to the conveying direction of the conveyor belt, the Y axis perpendicular to the conveying direction of the conveyor belt, and the coordinates of the visual recognition area are {(x1, x2), (y1, y2)};

[0015] Get the coordinate points within the interval time: Get the projection coordinate points in the XY coordinate system of the straight line where the vertical planes where the light beams emitted by any two on-state follow-spot lights intersect within the T time period;

[0016] If the Y coordinate values of all projection coordinate points are the same within the T time period, it is determined that the two tracking lights are tracking the same target metal can.

[0017] The logic for turning off the chase lights is that if the number of chase lights on both sides is the same, the chase lights on the left side of the conveyor belt are turned off. If the number of chase lights on both sides is different, the chase lights on the side with the higher number of chase lights are turned off. This avoids the situation where all chase lights on one side are turned on, but the deformed metal can still not cast light.

[0018] In order to remind the staff conveniently, the chasing lights are each individually connected to a buzzer, and the buzzer switches on and off synchronously with the chasing lights.

[0019] To avoid misjudgment, if the vertical planes of the lights of two turned-on follow-spot lights are parallel or overlap, they will not be tracked to identify the same target metal can at this moment.

[0020] As a preferred solution, if any of the acquired projection coordinates is outside the visual recognition area within the T time period, recognition is stopped. If it is outside the visual recognition area, recognition is not required, indicating that the same metal can is not being projected.

[0021] The beneficial effects of the invention are: the invention is a defective metal can guidance and indication system based on visual recognition. On the basis of the existing metal can stacking production line, a visual recognition camera is added. Through the use of the visual recognition camera, it can cooperate with the main control terminal to control the follow-up light to project light on the deformed metal can. After the projection light is turned off and projected onto the metal can, the staff can promptly remove it from the conveyor belt. In the above system, in order to avoid waste caused by multiple follow-up lights projecting on a metal can at the same time, a method is also proposed for turning off one of the follow-up lights projecting on the same metal can, and the follow-up light shutdown logic is limited by the set number and the open number, which can avoid the follow-up light being idle. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] By reading the detailed description of the preferred embodiment below, the solutions and advantages of the present application will become clear to those skilled in the art. The accompanying drawings are only for illustrating the preferred embodiment and are not to be considered as limiting the invention.

[0023] In the attached figure:

[0024] Figure 1 This is a schematic diagram of the top view of the structure of the invention;

[0025] Figure 2 A schematic diagram of the position state structure of the visual recognition area is provided for the invention;

[0026] Figure 3 A schematic diagram of the structure of a spotlight to indicate the state of a deformed metal can;

[0027] Figure 4 A schematic diagram for determining the status of multiple spotlights projecting onto the same metal can;

[0028] Figure 5 A schematic diagram for determining the status of multiple spotlights projecting onto the same metal can (with Figure 4 different moments);

[0029] Figure 6 This is a diagram indicating the status of a single follow-up light projecting onto the same metal can after it is turned off;

[0030] The components represented by the reference numerals in the figure are:

[0031] 1. Conveyor belt; 2. Air conveying mechanism; 3. Arrangement mechanism; 4. Lifting bracket; 5. Spotlight; 6. Metal can. DETAILED DESCRIPTION

[0032] like Figure 1-3 The shown system is a defective metal can guidance and indication system based on visual recognition, including a processing terminal and visual recognition cameras arranged on both sides of a conveyor belt 1. The visual recognition cameras are connected to the input end of the processing terminal and can transmit the captured images to the processing terminal. The number of visual recognition cameras is not less than two, so as to photograph and identify the metal cans from different positions, and to identify whether the metal cans are deformed or damaged from different positions. It should be noted that the visual recognition cameras are higher than the conveyor belt 1 and tilted downward to obtain a better field of view, and the visual recognition cameras are all set toward the conveyor belt 1, and a rectangular visual recognition area is set at the same position of the conveyor belt 1. The visual recognition cameras can take and identify photos of metal cans passing through the entire visual recognition area and identify whether they are deformed. Since this visual recognition method has been disclosed, the relevant recognition process will not be repeated. Only the core content of the present invention, namely, guiding and indicating deformed metal cans, will be introduced.

[0033] It should be noted that, since there may be multiple deformation problems in the metal cans passing through the visual recognition area at the same time, multiple chasing lights 5 are set under each visual recognition camera, and the chasing lights 5 can track the target metal can 6 based on the target metal can 6 identified by the visual recognition camera under the control of the processing terminal. In this way, the projection light of the chasing lights 5 moves synchronously with the target metal can 6 until the target metal can 6 exceeds the visual recognition area and is turned off. Therefore, when the projection light of the chasing lights 5 hits the target metal can, the staff can take the metal can out of the conveyor belt according to the above light indication information. And in order to facilitate the reminder of the staff, the chasing lights 5 are each separately connected to a buzzer, and the buzzer is switched on and off synchronously with the chasing lights 5. When the chasing lights 5 project light, the buzzer will also work and make a sound, thereby reminding the staff that there are deformed metal cans so that they can be quickly removed.

[0034] It should be noted that in the above structure, the length of the visual recognition area in the width direction of the conveyor belt 1 is not less than the width of the conveyor belt 1, that is, each metal can can be identified through the above visual recognition area. Due to the fact that the arrangement of the metal cans on the conveyor belt is not an array but random, in some cases, the metal cans may conflict with each other. Therefore, the above visual recognition area cannot identify metal cans with deformed conflicting surfaces. However, the probability of this happening is extremely low, so no special treatment is performed.

[0035] Therefore, based on the above structure, it can be achieved that, with the assistance of the spotlight 5, the staff can quickly find and identify the deformed metal can 6, and then the staff can identify and manually remove the metal can 6 according to the information of the spotlight 5. Since the number of deformed metal cans 6 on the entire conveyor belt 1 is not large, in this case, it is more appropriate to choose a lower-cost guidance and indication method than to set up a dedicated rejection structure. In this way, it can ensure that the staff are instructed at the first time while also completing the rejection of the deformed metal can 6. The manual rejection method is faster, and if the metal cans 6 in the surrounding positions are affected during the rejection process, the manual method can also quickly rearrange the metal cans 6 on the conveyor belt 1, which is more convenient.

[0036] Afterwards, based on the above structure, further, in order to facilitate human observation, the projection light color of the chasing light 5 is green, which is easier for the human eye to recognize and can form a sharp contrast with the metallic metal can, making it more obvious.

[0037] Finally, in order to better remove the metal cans, the specific method is that two visual recognition cameras are set, and are symmetrically arranged on both sides of the conveyor belt 1, and supported by a lifting bracket 4 fixed on the conveyor belt 1. First, since the metal cans block each other, the method of identifying from one side cannot complete the identification and indication of all metal cans. Therefore, by setting them separately and identifying from more positions, the effect is better.

[0038] As a preferred embodiment, Figure 4 、 5 As shown, in order to be able to project light onto multiple deformed metal cans 6 at the same time, the lifting bracket 4 is provided with multiple follow-up lights 5 spaced apart in the height direction below the visual recognition camera. Since a single follow-up light 5 can only identify and track a single metal can 6, if a large number of deformed metal cans 6 appear in a single time period, in order to ensure the best indication effect, a larger number of follow-up lights can be set to handle this situation, and the follow-up lights 5 can be rotated in the horizontal plane and the vertical plane, which is convenient for rotating and tracking the metal cans 6.

[0039] To avoid the situation where multiple chasing spots 5 project onto the same metal can 6, all chasing spots 5 are connected to the horizontal axis of rotation with an angle encoder. The angle encoder is connected to the processing terminal and can therefore provide real-time feedback of angle information. Therefore, if at least two chasing spots 5 are identified as tracking the same target metal can 6, the processing terminal can turn off at least one chasing spot 5, retain one chasing spot 5 for tracking, and eliminate the identification information of the target metal can 6 from the visual recognition camera corresponding to the turned-off chasing spot 5. This can be achieved by the above method. Only a single chasing spot 5 is required to indicate the staff. There is no enhancement effect when multiple chasing spots 5 are projected simultaneously, and increasing the number of chasing spots 5 used will affect the identification of subsequent metal cans 6.

[0040] In view of the above situation, the specific method for implementing the above identification is that the identification of at least two tracking lights 5 tracking the same target metal can 6 is specifically implemented through the following operations:

[0041] Establish a coordinate system: establish an XY coordinate system in the visual recognition area, with the X axis parallel to the conveying direction of conveyor belt 1, the Y axis perpendicular to the conveying direction of conveyor belt 1, and the coordinates of the visual recognition area are {(x1, x2), (y1, y2)};

[0042] Obtain the coordinate points within the interval time: obtain the projection coordinate points of the straight line where the vertical planes where the light beams emitted by any two on-state follow-spot lights 5 intersect in the XY coordinate system within the T time period;

[0043] If the Y coordinate values of all the projection coordinate points are the same within the T time period, it is determined that the two tracking lights 5 are tracking the same target metal can 6 .

[0044] By adopting the above-mentioned method, it is possible to avoid the situation where the angle encoders have the same values and are mistakenly believed to be tracking the same deformed metal can 6, resulting in one deformed metal can 6 not being indicated.

[0045] Afterwards, the logic for shutting down the chase lights is as follows: if the number of chase lights 5 on both sides is the same, the chase lights 5 on the left side of the conveyor belt 1 are shut down; if the number of chase lights 5 on both sides is different, the chase lights 5 on the side with the larger number of chase lights 5 are shut down. This prevents the situation where all chase lights 5 on one side are turned on, but the deformed metal can 6 still cannot cast light. It also prevents the situation where too many chase lights 5 on one side cause a deformed metal can 6 to appear, leaving no chase lights 5 tracking this location. If this situation occurs, the machine can be shut down.

[0046] In order to avoid misjudgment, if the vertical planes where the lights of the two turned-on follow-spot lights 5 are located are parallel to or overlap with each other, they will not be tracked to identify the same target metal can 6 at this moment.

[0047] Finally, if there is a coordinate point outside the visual recognition area among the obtained projection coordinate points within the T time period, the recognition is stopped. If it is outside the visual recognition area, no recognition is required, indicating that the same metal can 6 is not being projected.

Claims

1. A defective metal can guidance and indication system based on visual recognition, characterized in that: It comprises a processing terminal and visual recognition cameras arranged on both sides of a conveyor belt (1), and the number of the visual recognition cameras is not less than two, the visual recognition cameras are higher than the conveyor belt (1) and arranged to be tilted downward, the visual recognition cameras are all arranged toward the conveyor belt (1), and a rectangular visual recognition area is arranged at the same position of the conveyor belt (1); A plurality of tracking lights (5) are provided below each visual recognition camera. The tracking lights (5) can track the target metal can (6) based on the target metal can (6) recognized by the visual recognition camera under the control of the processing terminal, and the projection light of the tracking lights (5) moves synchronously with the target metal can (6) until the target metal can (6) exceeds the visual recognition area and then turns off; The length of the visual recognition area in the width direction of the conveyor belt (1) is not less than the width of the conveyor belt (1).

2. The defective metal can guidance and indication system based on visual recognition according to claim 1 is characterized in that: The color of the projection light of the chasing light (5) is green.

3. The defective metal can guidance and indication system based on visual recognition according to claim 1 is characterized in that: Two visual recognition cameras are provided and symmetrically arranged on both sides of the conveyor belt (1), and are supported by a lifting bracket (4) fixed to the conveyor belt (1).

4. The defective metal can guidance and indication system based on visual recognition according to claim 3 is characterized in that: The lifting bracket (4) is provided with a plurality of chasing lights (5) spaced apart in a height direction below the visual recognition camera, and the chasing lights (5) are capable of rotating on a horizontal plane and a vertical plane.

5. The defective metal can guidance and indication system based on visual recognition according to claim 4 is characterized in that: The rotation axes of all the chasing lights (5) on the horizontal plane are connected to angle encoders, and the angle encoders are communicatively connected to the processing terminal. If it is recognized that at least two chasing lights (5) are tracking the same target metal can (6), the processing terminal can turn off at least one chasing light (5), retain one chasing light (5) for tracking, and eliminate the identification information of the target metal can (6) of the visual recognition camera corresponding to the turned-off chasing light (5).

6. The defective metal can guidance and indication system based on visual recognition according to claim 5 is characterized in that: The identification of at least two tracking lights (5) tracking the same target metal can (6) is specifically achieved by the following operations: Establishing a coordinate system: establishing an XY coordinate system in the visual recognition area, with the X axis being parallel to the conveying direction of the conveyor belt (1), the Y axis being perpendicular to the conveying direction of the conveyor belt (1), and the coordinates of the visual recognition area being {(x1, x2), (y1, y2)}; Obtaining the coordinate points within the interval time: obtaining the projection coordinate points of the straight lines where the vertical planes where the light beams emitted by any two chase lights (5) in the turned-on state intersect in the XY coordinate system within the T time period; If the Y coordinate values of all the projection coordinate points are the same within the T time period, it is determined that the two tracking lights (5) are tracking the same target metal can (6).

7. The defective metal can guidance and indication system based on visual recognition according to claim 6 is characterized in that: When the number of chase lights (5) turned on at both sides is the same, the chase light (5) at the left side of the conveyor belt (1) is turned off. When the number of chase lights (5) turned on at both sides is different, the chase light (5) is turned off and the chase light (5) on the side with the larger number turned on is turned off.

8. The defective metal can guidance and indication system based on visual recognition according to claim 6 is characterized in that: The chasing lights (5) are each individually connected to a buzzer, and the buzzer and the chasing lights (5) are switched on and off synchronously.

9. The defective metal can guidance and indication system based on visual recognition according to claim 6 is characterized in that: If the vertical planes where the lights of the two turned-on tracking lights (5) are located are parallel to or overlap each other, then at this moment, they will not be tracked to identify the same target metal can (6).

10. The defective metal can guidance and indication system based on visual recognition according to claim 6, characterized in that: If, within the T time period, any of the acquired projection coordinate points is outside the visual recognition area, recognition is stopped.

Images