KR102458087B1 - Container photo damage reading of quay crane and labels of international maritime dangerous goods rule reading system - Google Patents

Abstract

The present invention relates to a container photo damage reading and international maritime dangerous goods rule reading system of a quay wall crane, and more specifically, when loading and unloading containers from a quay wall crane, it stores a picture of the container appearance and reads the damage, and the International Maritime Dangerous Goods Rule (IMDG) ) to a system for reading labels.
The present invention automatically detects whether the container is damaged by taking a picture of the container's appearance and determining whether the container is damaged through a damage reading algorithm, and notifies the center agent in real time to determine whether the container being worked on is damaged in real time. This has the advantage of reducing the number of disputes.
In addition, there is an advantage in that containers containing dangerous goods can be identified by reading the International Maritime Dangerous Goods Rule (IMDG) label through the International Maritime Dangerous Goods Rule (IMDG) reading algorithm.

Description

Container photo damage reading of quay crane and labels of international maritime dangerous goods rule reading system

The present invention relates to a container photo damage reading of a quay wall crane and a label reading system of the International Maritime Dangerous Goods Regulations. Specifically, the quay wall crane stores a picture of the container exterior and reads the damage when loading and unloading containers in the quay wall crane, and the container door position and door It relates to a system for reading open or not and reading the International Maritime Dangerous Goods (IMDG) label.

The container crane is installed on the quay wall of the wharf, and is a container dedicated crane for unloading containers from a container ship to the wharf and loading the containers on the pier to the ship. Container cranes are called variously, such as gantry crane (G/C), rail mounted quay crane (RMQC), or container or quay-side container crane. It is preferable to set it as a crane.

When a dispute occurs due to container damage, etc. during loading/unloading of containers in these container cranes, there is no clear evidence, so it is difficult to resolve the dispute, and the technology development necessary to secure the evidence is insufficient.

Korean Patent Registration Publication No. 10-2206662 (201.01.18) Korean Patent Registration Publication No. 10-2206658 (201.01.18)

An object of the present invention is to take a picture of the container exterior, determine whether the container is damaged through a damage reading algorithm, and read the International Maritime Dangerous Goods Rule (IMDG) label through the International Maritime Dangerous Goods Rule (IMDG) reading algorithm. will be.

In order to solve the above problems, the present invention provides a container photo collecting unit 100 for collecting container photos taken through a container recognition camera unit 60 that inspects the container by part, and the container photo collecting unit ( 100) to receive the container image from the image analysis unit 200 for analyzing the pattern, and the image analysis unit 200 to determine whether the container is damaged through the learning of the damage type of the container image whose pattern is analyzed through the image analysis unit 200 The damage determination unit 300 for recognizing the position of the container door through the container image analyzed by the pattern through the image analysis unit 200, and the door open determination unit 400 for determining whether the door is opened. Including, wherein the container photo collection unit 100 is a camera angle control unit 110 for adjusting the angle of the container recognition camera unit 60 to an angle capable of photographing each part of the container when the position of the container is confirmed ), to correct the distorted image among the container images collected through the container photo collecting unit 100 and improve the image quality, and to merge the container images taken through the camera unit close to the lane where the container is located into one image. and a container photo image correction unit 120, wherein the image analysis unit 200 includes an image pre-processing unit 210 for binarizing the container image provided from the container photo collecting unit 100, and the image pre-processing unit 210 ) through the image pattern analysis unit 220 for analyzing the pattern of the pre-processed container image, and the door open or not determining unit 400 is the container door rod in the container image provided from the container photo collection unit 100. A door pattern learning unit 410 for learning the end and vertical lines, and a door recognition unit 420 for determining whether the door is opened based on the information learned through the door pattern learning unit 410, Dangerous goods label recognition unit 500 for reading the label type based on the learned dangerous goods label among the container images taken by the container recognition camera unit 60 It provides a container photo damage reading and International Maritime Dangerous Goods Rules label reading system of the quay wall crane, characterized in that it includes.

The object of the present invention is to take a picture of the container exterior, determine whether the container is damaged through a damage reading algorithm, read the container door position and whether the door is open, and use the International Maritime Dangerous Goods Rule (IMDG) reading algorithm (IMDG) labels can be read.

1 is a block diagram showing the configuration of the present invention.
2 and 3 are views related to the container crane of the present invention.
4 is a view of the container photo collection unit 100.
5 and 6 are diagrams of the container photo image correction unit 110 of the container photo collection unit 100 .
7 and 8 are views of the damage determination unit 300 .
9 and 10 are views of the door recognition unit 400 .
11 and 12 are diagrams of the hazardous material label recognition unit 500 .

The terms or words used in the present specification and claims should not be construed as being limited to ordinary or dictionary meanings, and the inventor can properly define the concept of the term in order to best describe his invention. Based on the principle, it should be interpreted as meaning and concept consistent with the technical idea of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The present invention relates to a container photo damage reading of a quay crane and a label reading system of the International Maritime Dangerous Goods Regulations. Specifically, when loading and unloading containers from a quay wall crane, the container external photo is stored and the damage is read, and the International Maritime Dangerous Goods Rules ( IMDG) to a system for reading labels.

First, the crane 10, the container truck moving lane unit 20, the camera unit 30, the trolley position sensor 40, and the hoist position sensor 50 installed at the site of the present invention will be described.

The crane 10 is a trolley 11 installed at a location where the container truck moving lane part 20 is located and moving along a rail installed on a girder 11 made of a vertical member and a horizontal member, and the trolley 12 It is formed including a hoist 13 for lifting a heavy object, and a spreader 14 installed at the lower portion of the trolley 12 to grip and unload containers.

The girder 11 is made of a vertical member and a horizontal member, and the vertical member is made of four, and one first thread beam 11a is formed to connect the two vertical members formed in the direction facing the sea, and the first thread beam is formed. A second real beam 11b is formed at a position parallel to 11a, and the first portal beam 11c is connected to one end of the first real beam 11a and a vertical member connected to one end of the second real beam 11b, respectively. ) is formed, and a second portal beam 11d is formed to connect the other end of the first real beam 11a and the vertical member connected to the other end of the second real beam 11b, respectively.

The trolley 12 is to move along the rail installed on the girder.

The hoist 13 is for lifting the weight of the trolley 12 .

The spreader 14 is installed under the trolley 12 to grip and unload containers.

The container truck moving lane unit 20 is a ground on which the container truck 30 moves, and lanes through which a predetermined number of container trucks 30 move are formed.

The container truck 30 is provided with a YT license plate 31 on one side and the other side, respectively, and is preferably located on one side of the driver's side rearview mirror and one side of the top of the pillar at the passenger's side.

The container truck moving lane unit 20 may be divided into six lanes as shown in FIG. 2 , and the container truck 30 may be moved in each lane.

The trolley position recognition unit 40 is located on one side of the trolley 12 of the crane 10, but receives the position value of the trolley 12 and combines it with information about the container truck moving lane unit 20 for container recognition It is transmitted to the camera unit 60 .

The hoist and spreader position recognition unit 50 receives the position values of the hoist 13 and the spreader 14 of the crane 10 and transmits the transmitted position value to the container recognition camera unit 60 .

The hoist and spreader position recognition unit 50 recognizes the number when the position of the spreader 14 is located at a predetermined height from the ground (refer to the h mark in FIG. 3), and preferably recognizes the number when it is about 9M away. This is because, when the spreader 14 is located at a position of about 9M, the lane is generally determined to the left and the container type information is also determined.

The container recognition camera unit 60 is installed in a predetermined number of the portal beam and the real beam of the crane 10 to take a container number image, a YT number image, and a container door image.

The container recognition camera unit 60 includes a monitoring camera unit 61, a first portal beam camera unit 62, a second portal beam camera unit 63, a first real beam camera unit 64, and a second real beam camera. It is configured to include a portion 65 .

The monitoring camera unit 61 is located on one side of the upper side of the crane 10 and is not shown in the drawing to photograph the container truck moving lane unit 20 as a whole in order to monitor the container truck moving lane unit 20 as a whole. .

The first portal beam camera unit 62 is located on one side of the portal beam on one side of the crane 10, and preferably, two cameras are formed on one side and the other side of the first portal beam 11c of the girder 11. do.

The second portal beam camera unit 63 is positioned on one side of the portal beam on the other side of the crane 10, and preferably, two cameras are formed on one side and the other side of the second portal beam 11d of the girder 11. do.

The first real beam camera unit 64 is positioned on one side of the one side real beam of the crane 10 . Preferably, the camera is formed on one side of the second real beam 11b of the girder 11 .

The second true beam camera unit 65 is positioned on one side of the other side of the actual beam of the crane 10 , and the camera is preferably formed on one side of the first true beam 11a of the girder 11 .

4 is a view showing that the container truck 30 is taking a picture using the camera unit 60 for container recognition when passing the second lane, the container truck 30 is stopped at a specific location and the container 10 After recognizing the positions of the container truck 30 and the container 5 when the vehicle is seated, the angles of each camera are adjusted so that each part of the container truck 30 and the container 5 can be photographed. can be filmed

Referring to FIG. 1, the present invention includes a container photo collection unit 100, an image analysis unit 200, a damage determination unit 300, a door open determination unit 400, a hazardous material label recognition unit 500, a database unit ( 600) is included.

Each component is described in detail as follows.

Referring to Figure 4, the container photo collection unit 100 is located at a position set based on the position information of the trolley transmitted from the trolley position sensor 30 and the position information of the hoist transmitted from the hoist position sensor 40 When the container is located The angle of the container recognition camera unit 60 is adjusted to an angle capable of photographing each part of the container, and is for collecting container photos taken through the container recognition camera unit 60 whose angle is adjusted.

The container photo collection unit 100 is configured to include a camera angle control unit 110 and a container photo correction unit 120 .

The camera angle control unit 110 when the container is located at a position set based on the position information of the trolley transmitted from the trolley position sensor 30 and the position information of the hoist transmitted from the hoist position sensor 40, the container recognition camera unit The angle of (60) is to adjust the angle at which each part of the container can be photographed.

The container photo image correction unit 120 corrects a distorted image among the container images collected through the container photo collection unit 100 and improves the image quality, and converts the container image taken through the camera unit close to the lane where the container is located. for merging into an image of

FIG. 5(a) is a photographed container image showing a distorted image, and FIG. 5(b) is a view showing a state after correcting the distorted images and improving the image quality.

Fig. 6 (a) is a view showing a state in which one side of the container is separated into two parts and photographed, and Fig. 6 (b) is a view showing a state in which a container image photographed in two parts is merged into one image.

As such, the container photo image correction unit 120 corrects the overall size of the photographed structure image based on the distance to the container, which is the target structure, and the distance magnification of the camera lens for the photographed structure image of each area. At this time, the correction through the distance to the container and the distance magnification of the camera lens refers to adjusting the overall size of the photographed structure image based on the reference set value stored in advance for each distance and each distance magnification.

The correction made in the container photo image correction unit 120 is not to correct this distortion, but to simply increase the entire image based on the distance and distance magnification, so that only the size of the image is enlarged as it is in the distorted state. do.

As such, it is possible to correct the distorted part by increasing or reducing only the distorted area according to the reduced drawing of the actual structure by automatically joining it to the drawing of the reduced form of the actual structure based on the corrected structure image based on the reference setting value. .

The image analysis unit 200 is to receive the container image from the container photo collection unit 100 and analyze the pattern.

The image analysis unit 200 includes an image pre-processing unit 210 and an image pattern analysis unit 220 .

The image preprocessor 210 is for binarizing the container image provided from the container photo collection unit 100 .

The image pattern analysis unit 220 is for analyzing the pattern of the container image pre-processed through the image pre-processing unit 210 .

The image pattern analysis unit 220 analyzes the outline pattern of the object shape of the container image pre-processed through the image pre-processing unit 210. The image pre-processing unit 210 analyzes the container image taken as shown in FIG. After binarization through ), the outline pattern of the shape of an object was analyzed as shown in part k of FIG. 7(b).

The damage determination unit 300 is for determining whether the container is damaged by learning the damage form of the container image whose pattern is analyzed through the image analysis unit 200 .

It is possible to determine whether the container is damaged by checking whether the container shape is consistent with the previously learned container shape based on the information analyzed by the image analysis unit 200 to analyze the outer line pattern of the container shape.

Learning the container image in the form of damage is to learn the shape of the container (m1), the shape of the character and logo (m2), and the shape of the damage (m3), respectively, as shown in FIG. Logo and data location can be detected.

As a specific method for this, an irregular region (part m) may be detected in the binarized image after the corresponding image is binarized as shown in FIG. 7(b).

The door open determination unit 400 is for recognizing the position of the container door through the container image analyzed by the pattern through the image analysis unit 200 and determining whether the door is opened.

The door open determination unit 400 includes a door pattern learning unit 410 and a door recognition unit 420 .

The door pattern learning unit 410 is for learning the end of the container door rod and the vertical line from the container image provided from the container photo collection unit 100 .

Referring to FIG. 9, the door pattern learning unit 410 binarizes the image of the container door photographed as shown in FIG. 9(a) as shown in FIG. Thus, it is possible to determine the door and learn the bar end g1 and the vertical bar g2 of the container door as shown in FIG. 9(c).

The door recognition unit 420 is for determining whether the door is open based on the information learned through the door pattern learning unit 410 .

Referring to FIG. 10 , if it is determined that the container image provided in FIG. 10(a) is a container door, the end point and the middle link of each vertical bar are detected as shown in the red border of FIG. 10(b).

FIG. 10( c ) is a diagram illustrating a configuration of a detection area g3 of a door for determining whether a door is open including the result detected in 10( b ).

In FIG. 10( d ), it is possible to detect the straight line part g4 in the detection area g3 , and determine whether the door is opened by the distribution of the inclination histogram of the extracted straight line.

The dangerous goods label recognition unit 500 is for reading the label type based on the learned dangerous goods label among the container images photographed by the container recognition camera unit 60 .

11 is a diagram of the International Maritime Dangerous Goods Code (IMDG Code) dangerous goods label and mark, so that the type and shape of each label can be learned and read.

12 is a view separately showing only the hazardous material label part of the container image photographed by the container recognition camera unit 60, and classification and reading for the label type (n1 part) of FIG. 12 (a), and the symbol mark in the label (part n2) is read, and in the label form of FIG. 12(b), the type of label can be determined by recognizing the letter (n3 part) located in the upper middle of the label and recognizing the lower letter (part n4) in the label. have.

The size of the label photographed by the container recognition camera unit 60 is preferably 250 to 300 Pixel, and in order to recognize the label and the character, the vertical size of the character in the label must be at least 20 Pixel to be recognized as the character and the character It is possible to read the form classified as , but the scope is not limited.

The database unit 600 includes a container image corrected through the container photo collection unit 100 , a container image whose pattern is analyzed through the image analysis unit 200 , and a damage form learned through the damage determination unit 300 . Information on the label type and information on the label type read through the dangerous goods label recognition unit 500 is to be stored.

As such, the present invention can determine whether damage is caused through the photographed container image, and when damage occurs, the situation can be communicated to the worker by generating an alarm. There are advantages to be found.

The configuration shown in the embodiments and drawings described in the present specification as described above is only the most preferred embodiment of the present invention and does not represent all the technical spirit of the present invention, so various equivalents and modifications that can be substituted for them It should be understood that there may be examples.

5 container
10 crane
11 girder
12 trolley
13 hoist
14 spreader
20 Container truck moving lane
30 container truck

40 trolley position recognition unit
50 Hoist and spreader position recognition unit
60 Container recognition camera unit

100 Container Photo Collection Department
110 camera angle control unit
120 Container photo image correction unit
200 Image Analysis Department
210 image preprocessor
220 Image pattern analysis unit
300 damage judgement
400 Door open or not
410 Door Pattern Learning Department
420 door recognition unit
500 Dangerous goods label recognition unit
600 database division

Claims (4)

A container photo collection unit 100 for collecting container photos taken through a container recognition camera unit 60 that inspects the container for each part;
an image analysis unit 200 for receiving the container image from the container photo collection unit 100 and analyzing the pattern;
A damage determination unit 300 for determining whether the container is damaged by learning the damage form of the container image whose pattern is analyzed through the image analysis unit 200;
and a door open determination unit 400 for recognizing the position of the container door through the container image whose pattern is analyzed through the image analysis unit 200 and determining whether the door is opened,
The container photo collection unit 100 is
When the position of the container is confirmed, the camera angle control unit 110 for adjusting the angle of the container recognition camera unit 60 to an angle at which each part of the container can be photographed;
A container photo image for correcting a distorted image among the container images collected through the container photo collecting unit 100 and improving the image quality, and merging the container images taken through the camera unit close to the lane where the container is located into one image Comprising a correction unit 120,
The image analysis unit 200
An image pre-processing unit 210 for binarizing the container image provided from the container photo collection unit 100;
and an image pattern analysis unit 220 for analyzing the pattern of the container image pre-processed through the image pre-processing unit 210,
The door open determination unit 400 is
a door pattern learning unit 410 for learning the end and vertical lines of the container door rods from the container image provided from the container photo collection unit 100;
Container photo damage reading system of a quay wall crane, characterized in that it comprises a door recognition unit 420 for determining whether the door is open based on the information learned through the door pattern learning unit 410. A container photo collection unit 100 for collecting container photos taken through a container recognition camera unit 60 that inspects the container for each part;
an image analysis unit 200 for receiving the container image from the container photo collection unit 100 and analyzing the pattern;
A damage determination unit 300 for determining whether the container is damaged by learning the damage form of the container image whose pattern is analyzed through the image analysis unit 200;
A door open determination unit 400 for recognizing the position of the container door through the container image analyzed by the pattern through the image analysis unit 200 and determining whether the door is opened;
and a dangerous goods label recognition unit 500 for reading the label type based on the learned dangerous goods label among the container images taken by the container recognition camera unit 60,

The container photo collection unit 100 is
When the position of the container is confirmed, the camera angle control unit 110 for adjusting the angle of the container recognition camera unit 60 to an angle at which each part of the container can be photographed;
A container photo image for correcting a distorted image among the container images collected through the container photo collecting unit 100 and improving the image quality, and merging the container images taken through the camera unit close to the lane where the container is located into one image Comprising a correction unit 120,
The image analysis unit 200
An image pre-processing unit 210 for binarizing the container image provided from the container photo collection unit 100;
and an image pattern analysis unit 220 for analyzing the pattern of the container image pre-processed through the image pre-processing unit 210,
The door open determination unit 400 is
a door pattern learning unit 410 for learning the end and vertical lines of the container door rods from the container image provided from the container photo collection unit 100;
Container photo damage reading and international maritime dangerous goods rule label reading of a quay wall crane, characterized in that it includes a door recognition unit 420 for determining whether a door is opened based on the information learned through the door pattern learning unit 410 system. delete delete

Images