Ship non-consumable intelligent counting method and system based on image recognition
Through the intelligent counting method based on image recognition, the problem of non-consumables measurement during shipbuilding is solved, efficient and accurate measurement is achieved, and manpower investment is reduced.
Patent Information
- Application Number
- CN202510226336.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-27
- Publication Date
- 2025-06-13
AI Technical Summary
During the shipbuilding process, the measurement of non-consumable products such as steel pipes, fasteners, screws, etc. is difficult to measure, resulting in a lot of manpower consumption, and the existing technology is difficult to effectively solve this problem.
Using an intelligent counting method based on image recognition, the camera collects the photo of the item, performs secondary processing and class library comparison, and selects a suitable calculation method to identify the specific quantity or total volume of the item.
It realizes efficient measurement of workpieces during shipbuilding, reduces manpower investment, and improves measurement accuracy and efficiency.
Smart Images
Figure CN120147598A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of AI intelligent recognition technology, and in particular to an intelligent counting method and system for non-consumable items of ships based on image recognition. Background Art
[0002] The image recognition method for intelligent counting of non-consumable ship parts is an innovative way to count ship-related tooling parts. It uses modern scientific and technological means, such as intelligent sensing equipment, big data analysis and artificial intelligence technology, to integrate the recognition and counting methods of object shapes. It has a high ability to effectively recognize and count tooling parts and other equipment in the shipbuilding industry, which can save a lot of counting manpower and effectively count their quantity.
[0003] In the intelligent counting of image recognition, there are many test items, such as steel pipes, fasteners, screws, wires, scaffolding boards of different lengths, connecting strips, etc. Therefore, it is particularly important to accurately determine the category of the items and the counting method.
[0004] Existing technologies include 3D camera-based recognition technology and deep learning-based recognition technology. They use image processing and feature extraction in artificial intelligence and then confirm the classification of items after processing by a classifier. However, it is difficult to solve the measurement problem of tooling parts in the shipbuilding process. For example, consumables do not need to be counted after they are used and worn out. However, non-consumables, such as steel pipes, fasteners, screws, pins, scaffolding boards of different lengths, connecting strips, etc., are difficult to count after recycling, which consumes a lot of manpower.
[0005] Therefore, a method and system for intelligent counting of non-consumable items of ships based on image recognition are provided. Summary of the invention
[0006] The purpose of the present invention is to overcome the existing defects and provide an intelligent counting method for non-consumable parts of ships based on image recognition, which solves the measurement problem of tooling parts in the shipbuilding process.
[0007] The technical solution to achieve the above purpose is:
[0008] The present invention provides an intelligent counting method for non-consumable items of ships based on image recognition, comprising:
[0009] Step S1, taking a photo of a designated location through a camera or directly taking a photo, collecting data, and storing it;
[0010] Step S2, after storing the photo, it is processed again and compared with the category library in the internal library to obtain the compared items;
[0011] Step S3, after comparing the items, select a calculation method suitable for the items;
[0012] Step S4, identify the key information required for calculation by selecting an appropriate calculation method, and finally obtain the specific quantity or total volume of the item;
[0013] Step S5, return the stored image of the comparison result of the library, the original input photo, and the specific data obtained.
[0014] Preferably, in step S1, the photos collected by the camera or taking pictures are divided into pictures in three directions: front, side, and top.
[0015] Preferably, in step S3, there are two calculation methods in total:
[0016] The first one is for items that are easy to count: perform feature extraction on the extracted image through image processing, combine them into feature vectors, and compare with the single data in the library to directly obtain the specific quantity;
[0017] The second one is for items that are difficult to count: use the smallest cube or cuboid to frame a single part for the extracted image through image processing, calculate the volume, then divide the volume of the container recognized by the picture by this data to get the minimum number of parts that can be installed under this capacity, and then directly measure the volume of a single part and use the same method to find the maximum part capacity, and take the average of the two as the estimated value of the part.
[0018] Preferably, in step S4, if the first calculation method for items that are easy to count is selected, the key information required includes but is not limited to the length of the round opening surface of the corresponding item to obtain the specific quantity;
[0019] If the second calculation method for items that are difficult to count is selected, the key information required includes but is not limited to length, width, and height to calculate the total volume.
[0020] An intelligent counting system for non-consumable items on ships based on image recognition according to the second aspect of the present invention includes:
[0021] A data acquisition module, used to transmit the photo to the specified position through the camera or directly taking pictures for data acquisition;
[0022] A storage module, used to store the collected photos;
[0023] A data processing module, used to perform secondary processing on the stored photos and conduct library comparison in the internal library;
[0024] A selection module, used to select an appropriate calculation method through the comparison result of the library;
[0025] An identification module, used to identify the key information required for calculation by selecting an appropriate calculation method, and finally obtain the specific quantity or total volume of the item;
[0026] A result return module for returning the stored image of the comparison result of the class library, the original input photo, and the specific data obtained.
[0027] Preferably, in the data acquisition module, the photos collected by the camera or taking pictures are divided into pictures in three directions: front, side, and top.
[0028] Preferably, in the selection module, there are two calculation methods in total:
[0029] The first one is for items that are easy to count: Through image processing, feature extraction is performed on the extracted image and combined into a feature vector to compare with the single data in the class library to directly obtain the specific quantity;
[0030] The second one is for items that are difficult to count: Through image processing, the extracted image is framed by the smallest cube or cuboid for each single part, the volume is calculated, and then the volume of the container recognized by the picture is divided by this data to obtain the minimum number of parts that can be loaded under this capacity. Then, the volume of a single part is directly measured and the maximum part capacity is obtained in the same way, and the average of the two is taken as the estimated value of this part.
[0031] Preferably, in the recognition module, if the first method is selected, that is, the calculation method for items that are easy to count, the required key information includes but is not limited to the length of the round opening surface of the corresponding item to obtain the specific quantity;
[0032] If the second method is selected, that is, the calculation method for items that are difficult to count, the required key information includes but is not limited to length, width, and height to calculate the total volume.
[0033] The beneficial effects of the present invention are: Through secondary processing, the present invention conducts a comparison of the class library in the internal library, analyzes specific items, then selects a suitable calculation method based on the items, and further obtains the key information of the items. Finally, the specific quantity or total volume of the items is calculated through this key information, effectively solving the measurement problem of tooling parts during the shipbuilding process. Description of the Drawings
[0034] Figure 1 is a flowchart of an intelligent counting method for non-consumable items of ships based on image recognition according to the present invention;
[0035] Figure 2 is a module diagram of an intelligent counting system for non-consumable items of ships based on image recognition according to the present invention. Detailed Embodiments
[0036] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. In the description of the present invention, it should be noted that the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings, which are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as a limitation on the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only and cannot be understood as indicating or implying relative importance.
[0037] The present invention will be further described below in conjunction with the accompanying drawings.
[0038] like Figure 1 As shown, a method for intelligent counting of non-consumable items of a ship based on image recognition comprises:
[0039] Step S1, input the designated position photo through the camera or directly take a photo, collect data, and store it.
[0040] In the embodiment, the photos collected by the camera or the camera are divided into three directions: front, side, and top.
[0041] Step S2, after storing the photos, they are processed again and compared with the internal library to obtain the compared items; for example, steel pipes, fasteners, screws, pins, scaffolding boards of different lengths, connecting strips, etc.
[0042] Step S3, after comparing the items, select a calculation method suitable for the items.
[0043] In the embodiment, there are two calculation methods:
[0044] The first one is for items that are easy to count: feature extraction is performed on the extracted image through image processing and combined into feature vectors to compare with single data in the library to directly obtain the specific quantity; for example, steel pipes.
[0045] The second method is for items that are difficult to count: through image processing, the extracted image is framed with a minimum cube or rectangular block to determine the volume. The volume of the container identified by the image is then divided by the data to determine the minimum number of parts that can be loaded under the capacity. The volume of a single part is then directly measured and the maximum part capacity is determined in the same way. The average of the two is the estimated value of the part; for example, small objects such as fasteners and pins.
[0046] Step S4, identifying the key information required for calculation by selecting an appropriate calculation method, and finally obtaining the specific quantity or total volume of the item.
[0047] In the embodiment, if the first method is selected, which is the calculation method for items that are easy to count, the key information required includes, but is not limited to, the length of the round opening surface of the corresponding item, to obtain the specific quantity. For example, for steel pipes: the length of the round opening surface of the pipe stack is used to determine how many steel pipes of a certain length there are in total.
[0048] If the second method is selected, which is the calculation method for items that are difficult to count, the key information required includes, but is not limited to, length, width, and height, to calculate the total volume. For example, for pins: identify their length, width, and height, and calculate the total volume.
[0049] Step S5: Return the saved image of the class library comparison result, the original input photo, and the obtained specific data.
[0050] As Figure 2 shown, an intelligent counting system for non-consumable items on ships based on image recognition includes: a data acquisition module 1, a storage module 2, a data processing module 3, a selection module 4, an identification module 5, and a result return module 6.
[0051] The data acquisition module 1 is used to transmit the photo to the specified position through the camera or by directly taking a photo for data acquisition.
[0052] In the embodiment, the photos collected through the camera or by taking a photo are divided into pictures in three directions: the front, the side, and the top.
[0053] The storage module 2 is used to store the collected photos.
[0054] The data processing module 3 is used to perform secondary processing on the stored photos and conduct a class library comparison in the internal library.
[0055] The selection module 4 is used to select a suitable calculation method based on the class library comparison result.
[0056] In the embodiment, there are two calculation methods in total:
[0057] The first one is for items that are easy to count: Through image processing, feature extraction is performed on the extracted image, and the feature vectors are combined and compared with the single data in the class library to directly obtain the specific quantity.
[0058] The second one is for items that are difficult to count: Through image processing, the extracted image is framed by the smallest cube or cuboid for each single part, the volume is calculated, and then the volume of the container recognized by the picture is divided by this data to obtain the minimum number of parts that can be loaded under this capacity. Then, the volume of a single part is directly measured, and the maximum part capacity is obtained in the same way. Take the average of the two as the estimated value of this part.
[0059] The identification module 5 is used to identify the key information required for calculation by selecting a suitable calculation method, and finally obtain the specific quantity or total volume of the item.
[0060] In the embodiment, if the first method, which is a calculation method for easily countable items, is selected, the key information required includes, but is not limited to, the length of the round opening surface of the corresponding item, to obtain the specific quantity;
[0061] If the second method, which is a calculation method for difficult-to-count items, is selected, the key information required includes, but is not limited to, length, width, and height, to calculate the total volume.
[0062] The result return module 6 is used to return the stored image of the class library comparison result, the original input photo, and the obtained specific data.
[0063] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements on some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the scope of the technical solutions of the embodiments of the present invention.
Claims
1. An intelligent counting method for non-consumable items of ships based on image recognition, characterized in that: include: Step S1, taking a photo of a designated location through a camera or directly taking a photo, collecting data, and storing it; Step S2, after storing the photo, it is processed again and compared with the category library in the internal library to obtain the compared items; Step S3, after comparing the items, select a calculation method suitable for the items; Step S4, identifying the key information required for calculation by selecting an appropriate calculation method, and finally obtaining the specific quantity or total volume of the item; Step S5, returning the stored image of the library comparison result, the original input photo, and the specific data obtained.
2. The method for intelligent counting of non-consumable items of ships based on image recognition according to claim 1 is characterized in that: In step S1, the photos collected by the camera or the camera are divided into three directions: front, side and top.
3. The method for intelligent counting of non-consumable items of ships based on image recognition according to claim 1 is characterized in that: In step S3, there are two calculation methods: The first one is for items that are easy to count: the extracted images are processed to extract features and combined into feature vectors to compare with the single data in the library to directly obtain the specific quantity; The second method is for items that are difficult to count: through image processing, the extracted image is framed with a minimum cube or rectangular block to obtain the volume. The volume of the container recognized by the image is then divided by the data to obtain the minimum number of parts that can be loaded under the capacity. The volume of a single part is then directly measured and the maximum part capacity is calculated in the same way. The average of the two is the estimated value of the part.
4. The method for intelligent counting of non-consumable items of ships based on image recognition according to claim 1 is characterized in that: In step S4, if the first method is selected, which is a calculation method for objects that are easy to count, the key information required includes but is not limited to the length of the round mouth surface of the corresponding object to obtain the specific quantity; If you choose the second method, which is for calculating items that are difficult to count, the key information required includes but is not limited to length, width, height, and total volume.
5. The ship non-consumable intelligent counting system based on image recognition is characterized by: include: The data collection module is used to collect data by transmitting photos of designated locations through a camera or by taking photos directly; A storage module, used for storing collected photos; The data processing module is used to perform secondary processing on the stored photos and compare the categories in the internal library; Selection module, used to select the appropriate calculation method through library comparison results; An identification module is used to identify the key information required for calculation by selecting an appropriate calculation method, and finally obtain the specific quantity or total volume of the item; The result return module is used to return the stored image of the library comparison result, the original input photo, and the specific data obtained.
6. The ship non-consumable parts intelligent counting system based on image recognition according to claim 5 is characterized in that: In the data collection module, the photos collected by the camera or the camera are divided into three directions: front, side and top.
7. The ship non-consumable parts intelligent counting system based on image recognition according to claim 5 is characterized in that: In the selection module, there are two calculation methods: The first one is for items that are easy to count: the extracted images are processed to extract features and combined into feature vectors to compare with the single data in the library to directly obtain the specific quantity; The second method is for items that are difficult to count: through image processing, the extracted image is framed with a minimum cube or rectangular block to obtain the volume. The volume of the container recognized by the image is then divided by the data to obtain the minimum number of parts that can be loaded under the capacity. The volume of a single part is then directly measured and the maximum part capacity is calculated in the same way. The average of the two is the estimated value of the part.
8. The ship non-consumable parts intelligent counting system based on image recognition according to claim 5 is characterized in that: In the identification module, if the first method is selected, which is a calculation method for objects that are easy to count, the key information required includes but is not limited to the length of the round mouth surface of the corresponding object to obtain the specific quantity; If you choose the second method, which is for calculating items that are difficult to count, the key information required includes but is not limited to length, width, height, and total volume.