Automatic agilawood separating and processing system and method based on nuclear magnetic resonance imaging

Abstract

The invention relates to the technical field of agilawood processing, and provides an automatic agilawood separating and processing system and method based on nuclear magnetic resonance imaging. The system comprises a conveying belt, a positioning table, an MRI scanning device, an image processing unit, a rough cutting device, a fine carving device and a control system. The conveying belt is used for conveying agilawood to the positioning table, and the positioning table is responsible for fixing the agilawood and providing accurate positioning. And the MRI scanning equipment is arranged along the moving path of the conveying belt and the positioning table and is used for acquiring a high-resolution three-dimensional image of the agilawood inside the agilawood log. And the image processing unit optimizes the three-dimensional model according to the MRI image, accurately distinguishes the positions and structures of the agilawood and the peripheral white wood, generates a two-stage processing path, and realizes efficient and accurate separation of the agilawood and the peripheral white wood. According to the system, the two-stage processing flow of rough cutting and fine carving is introduced, so that the processing precision and efficiency of agilawood are improved, and meanwhile, the labor cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of agarwood processing, in particular to an automated separation and fine processing system for agarwood, which utilizes nuclear magnetic resonance imaging (MRI) technology and artificial intelligence image processing algorithms to optimize the processing efficiency and accuracy of agarwood. Background Art

[0002] Agarwood is widely used in spices and medicinal materials due to its unique aroma and medicinal value. Traditional agarwood processing mainly relies on manual operations, including cutting and carving with manual tools. This method is not only inefficient but also prone to resource waste, as it is difficult to precisely control the cutting depth and shape during manual processing, resulting in the loss of precious agarwood materials.

[0003] In addition, existing automated agarwood processing technologies such as CT scanning and automatic carving systems, although improving the efficiency and accuracy of manual operations, still have some limitations. For example, CT scanners in the existing technology are mainly used to obtain two-dimensional or low-resolution three-dimensional images of agarwood, which limits the accuracy of model reconstruction and the optimization of cutting paths. Moreover, these systems usually only include a single-stage carving process without distinguishing between rough cutting and fine carving steps, resulting in suboptimal processing efficiency and material utilization rate. Therefore, there is an urgent need for a two-stage separation and processing technology based on high-resolution three-dimensional images to improve the efficiency and reliability of agarwood automated separation technology.

[0004] Chinese Patent Document CN117863296A discloses an "Automated Agarwood Separation System and Method", which adopts a conveyor line, a rotary table, a CT scanner, a carving component and a control main board; the CT scanner is arranged close to the conveyor line to obtain CT slice images of agarwood; the rotary table, the CT scanner and the carving machine are all electrically connected to the control main board, and the control main board is configured to reconstruct the 3D models of agarwood and white wood according to the CT slice images, and then generate carving paths according to the 3D models. The above solution mainly uses a CT scanner to obtain two-dimensional or low-resolution three-dimensional images of agarwood, which limits the accuracy of model reconstruction and the optimization of cutting paths. Moreover, these systems usually only include a single-stage carving process without distinguishing between rough cutting and fine carving steps, resulting in suboptimal processing efficiency and material utilization rate. Summary of the Invention

[0005] The object of the present invention is to provide an automated separation and processing system and method for agarwood based on nuclear magnetic resonance imaging in view of the deficiencies of the existing technology.

[0006] Technical solution: The technical solution adopted by the present invention to solve the problem is: An agarwood automatic separation and processing system based on nuclear magnetic resonance imaging, and the separation and processing include: a conveying module; a fixing module; a nuclear magnetic resonance scanning device; an image processing and control module; a rough processing module; a fine processing module; a control module; the conveying module includes a conveyor belt; the fixing module includes a positioning table fixed on the conveyor, and an agarwood fixture fixed on the positioning table fixes a single piece of wood; the nuclear magnetic resonance scanning device surrounds the conveyor belt; the image processing and control module is located on the side of the conveyor belt; the rough processing module includes rough processing tool equipment fixed inside the conveyor belt; the fine processing module includes a multi-axis engraving machine fixed on the side of the conveyor belt.

[0007] Preferably, the conveyor belt has multiple independently adjustable speeds, and the whole is made of non-metallic materials and non-ferromagnetic metal materials. The speed of the conveyor belt passing through the nuclear magnetic resonance scanning device matches the scanning speed of the nuclear magnetic resonance scanning device, and the speed of the conveyor belt passing through the rough processing tool equipment is adjusted according to the cutting speed.

[0008] Preferably, the positioning table and the agarwood fixture are made of non-metallic materials and non-ferromagnetic metal materials throughout. The positioning table can move in the vertical direction of the conveyor belt operation and can rotate. The agarwood fixture is composed of several groups of fixtures, and the materials used have different imaging from the wood in the nuclear magnetic resonance scanning device.

[0009] Preferably, during the processing stage of the multi-axis engraving machine for the wood clamped by the positioning table, the wood is rotated to a direction perpendicular to the ground by rotating the fixed table.

[0010] Preferably, the movement paths of the rough processing tool equipment, the multi-axis engraving machine and the positioning table are all automatically generated by the image processing and control module.

[0011] The present invention also provides an agarwood automatic separation and processing method based on nuclear magnetic resonance imaging, which specifically includes the following steps:

[0012] S1: Use a nuclear magnetic resonance scanning device to obtain a high-resolution three-dimensional stereoscopic image of the agarwood.

[0013] S2: Input the obtained high-resolution three-dimensional stereoscopic image into the image processing and control module for the first fitting, and initially fit the three-dimensional image into multiple planes perpendicular to the running direction of the conveyor belt.

[0014] S3: Second fitting, perform finer-grained processing, and further fit it into a polyhedron composed of multiple smaller planes on the basis of the first fitting to ensure high-precision surface matching.

[0015] S4: Automatically generate the combined motion path of the rough machining tool equipment and the fine machining multi-axis engraving machine according to the polyhedron model after two fittings, so as to achieve efficient and precise processing of agarwood.

[0016] S5: Execute the generated path above, perform preliminary shape rough machining through the rough machining tool equipment, and then use the multi-axis engraving machine for detailed fine machining to complete the final processing of agarwood.

[0017] Preferably, the first step of fitting the three-dimensional stereoscopic image of high-resolution agarwood in S2 includes: First, segment the internal agarwood at a fixed equal interval, then use the maximum intensity projection technology for the three-dimensional data of each small section of agarwood log, project the three-dimensional image data onto the chordal plane of the agarwood log, and find the maximum cross-section; Then calculate the minimum circumscribed rectangle enclosing a combination of multiple maximum cross-sections in three-dimensional space, and use the minimum circumscribed rectangle as the cutting plane.

[0018] Preferably, the second fitting in S3 for finer-grained processing includes the steps of: First, denoise the high-resolution three-dimensional picture; Then use the three-dimensional convex hull algorithm to generate an initial polyhedron enclosing the agarwood; Then use the mesh deformation algorithm combined with the gradient descent method to iteratively optimize the vertex positions of the polyhedron to make it more accurately fit the surface shape of the agarwood.

[0019] Preferably, in S4, according to the polyhedron model after two fittings, the polyhedron model of the second fitting is inside the polyhedron model of the first fitting, and the outer surface of the log is based on the polyhedron after the first fitting.

[0020] Preferably, after automatically generating the combined motion path of the rough machining tool equipment and the fine machining multi-axis engraving machine in S4, automatically calculate the processing time of the rough machining tool equipment and the fine machining multi-axis engraving machine, and the image processing and control module automatically adjusts the scanning speed and resolution of the nuclear magnetic resonance scanning equipment.

[0021] Beneficial effects: (1) The present invention obtains a high-resolution three-dimensional stereoscopic image through a nuclear magnetic resonance scanning device, and combines a two-stage fitting algorithm to reduce the error rate of three-dimensional reconstruction of the agarwood structure; (2) The present invention adopts a collaborative mechanism of rough machining and fine machining to improve the overall processing efficiency; (3) The image processing and control module of the present invention has intelligent computing capabilities, can dynamically adjust nuclear magnetic resonance scanning parameters, conveyor belt speed and cooperate with processing equipment to maximize the utilization rate of system resources; (4) The present invention adopts a polyhedron fitting algorithm iteratively optimized by the gradient descent method to improve the extraction rate of effective components of agarwood. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the overall structure of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0023] The present invention will be further illustrated below in conjunction with the accompanying drawings and specific embodiments. These embodiments are implemented on the premise of the technical solution of the present invention. It should be understood that these embodiments are only used to illustrate the present invention and not to limit the scope of the present invention.

[0024] As Figure 1 shown, an automated separation and processing system for agarwood based on nuclear magnetic resonance imaging includes: a conveying module; a fixing module; a nuclear magnetic resonance scanning device 4; an image processing and control module; a rough processing module; a fine processing module; a control module; the conveying module includes a conveyor belt 1; the fixing module includes a positioning table 3 fixed to the conveyor 1, and an agarwood fixture 7 fixed to the positioning table 3 fixes a single piece of wood 8; the nuclear magnetic resonance scanning device 4 surrounds the conveyor belt 1; the image processing and control module 2 is located on the side of the conveyor belt; the rough processing module includes a rough processing tool device 5 fixed inside the conveyor belt 1; the fine processing module includes a multi-axis engraving machine 6 fixed on the side of the conveyor belt.

[0025] The conveyor belt 1 has multiple independently adjustable speeds. The speed of the section passing through the nuclear magnetic resonance scanning device 4 matches the scanning speed of the nuclear magnetic resonance scanning device 4. The speed of the section passing through the rough processing tool device 5 is adjusted according to the cutting speed. The multi-axis engraving machine 6 can remain stationary during its section. The entire conveyor belt is made of non-metallic materials and non-ferromagnetic metal materials, which are made of rubber tracks and plastic brackets respectively.

[0026] Before the agarwood 8 is fixed to the positioning table 3 and the agarwood fixture 7, a groove with a longitudinal length of 5 cm, a radial length of 1 cm, and a width of 1 cm is cut at one end of the agarwood 8 according to the clamping position of the agarwood fixture 7 for matching and clamping the agarwood fixture 7.

[0027] The positioning table 3 and the agarwood fixture 7 are entirely made of non-metallic materials and non-ferromagnetic metal materials. The agarwood fixture 7 is composed of 4 groups of asymmetric fixtures, and the materials used have different images from the wood in the nuclear magnetic resonance scanning device 4, and can establish a coordinate axis based on the asymmetric groove in the three-dimensional imaging.

[0028] The agarwood 8 is fixed to the positioning table 3 and the agarwood fixture 7, and a high-resolution three-dimensional stereoscopic image of the agarwood is obtained using the nuclear magnetic resonance scanning device. The three-dimensional stereoscopic image is sent to the image processing and control module 2.

[0029] The high-resolution three-dimensional stereoscopic image of the agarwood is first fitted. First, the maximum cross-section of the agarwood is found at a fixed equal distance, the gradient threshold of the edge of the maximum cross-section of the agarwood is calculated, the contour of the maximum cross-section is adjusted according to the set threshold, and finally the minimum circumscribed rectangle of the maximum cross-section contour is used as the cutting surface. According to the cutting surface information, the movement paths of the positioning table 3 and the agarwood fixture 7 and the movement trajectory of the conveyor belt where they are located are automatically generated, and the processing time is automatically calculated.

[0030] The positioning table 3 and the agarwood fixture 7 move to the rough machining tool equipment section 5. The rough machining tool equipment 5 starts to operate. The positioning table 3 and the agarwood fixture 7 move to the multi-axis engraving machine section 6, and the rough machining tool equipment 5 stops operating.

[0031] The positioning table 3 can move in the vertical direction along the conveyor belt and can rotate. According to the first fitted plane, the positioning table 3 moves in the vertical direction of the conveyor belt to the specified position, then moves forward along the conveyor belt direction, and then returns to the original position and reciprocates to complete the rough machining module.

[0032] Based on the rough machining and the first fitting, the image processing and control module 2 will perform the second fitting for a finer-grained processing. The polyhedron model of the second fitting is inside the polyhedron model of the first fitting. The outer surface of the log is based on the polyhedron after the first fitting. First, noise reduction is performed on the high-resolution three-dimensional image, and a polyhedron that can enclose the entire agarwood is initialized. Then, surface adjustment is carried out, the gradient of the fitting error is calculated for each face, and the gradient descent method is applied to adjust the vertex positions of the polyhedron to generate the processing plan and processing time for the second multi-axis engraving machine 6.

[0033] During the processing stage of the multi-axis engraving machine 6 for the wood 8 clamped by the positioning table 3, the wood 8 is rotated to a direction perpendicular to the ground by rotating the fixed table, and then the multi-axis engraving machine 6 engraves from top to bottom according to the engraving path.

[0034] The image processing and control module 2 automatically adjusts the scanning speed and resolution of the nuclear magnetic resonance scanning equipment 4 according to the scanning time of the nuclear magnetic resonance scanning equipment 4, the rough machining time of the rough machining tool equipment 5, and the processing time of the multi-axis engraving machine 6.

[0035] The automated separation and processing method of agarwood based on nuclear magnetic resonance imaging of the above equipment specifically includes the following steps:

[0036] S1: Use the nuclear magnetic resonance scanning equipment to obtain a high-resolution three-dimensional stereoscopic image of the agarwood.

[0037] S2: Input the obtained high-resolution three-dimensional stereoscopic image into the image processing and control module for the first fitting, and initially fit the three-dimensional image into multiple planes perpendicular to the running direction of the conveyor belt.

[0038] S3: Perform the second fitting for a finer-grained processing, and further fit it into a polyhedron composed of multiple smaller planes on the basis of the first fitting to ensure high-precision surface matching.

[0039] S4: According to the polyhedron models after the two fittings, automatically generate the combined movement paths of the rough machining tool equipment and the fine machining multi-axis engraving machine to achieve the efficient and precise processing of the agarwood.

[0040] S5: Execute the generated path above, conduct preliminary rough machining of the initial shape through a rough machining tool device, and then use a multi-axis engraving machine for detailed finishing to complete the final processing of the agarwood.

[0041] The first step of fitting the high-resolution three-dimensional stereo image of agarwood in S2 above includes: First, segment the internal agarwood at a fixed equal interval, then use the maximum intensity projection technique for the three-dimensional data of each small segment of agarwood log, project the three-dimensional image data onto the tangential plane of the agarwood log, and find the maximum cross-section; then calculate the minimum circumscribed rectangle that encloses a combination of multiple maximum cross-sections in three-dimensional space, and use the minimum circumscribed rectangle as the cutting surface.

[0042] The second fitting in S3 above for finer-grained processing includes the steps: First, denoise the high-resolution three-dimensional image; then use the three-dimensional convex hull algorithm to generate an initial polyhedron that encloses the agarwood; then use the mesh deformation algorithm combined with the gradient descent method to iteratively optimize the vertex positions of the polyhedron to make it more accurately fit the surface shape of the agarwood.

[0043] In S4 above, according to the polyhedron models after the two fittings, the polyhedron model of the second fitting is inside the polyhedron model of the first fitting, and the outer surface of the log is based on the polyhedron after the first fitting.

[0044] After automatically generating the combined motion path of the rough machining tool device and the multi-axis engraving machine for fine machining in S4 above, automatically calculate the machining times of the rough machining tool device and the multi-axis engraving machine for fine machining, and automatically adjust the scanning speed and resolution of the nuclear magnetic resonance scanning device by the image processing and control module.

[0045] The above specific implementation manner is only a preferred embodiment of the present invention and is not used to limit the implementation and the scope of the claims of the present invention. All equivalent changes and modifications made in accordance with the content of the patent protection scope of the present invention application shall be included within the scope of the present invention patent application.

Claims

1. An automated agarwood separation and processing system based on nuclear magnetic resonance imaging, characterized in that: include: A conveying module, a fixing module, a nuclear magnetic resonance scanning device (4), an image processing and control module, a rough processing module, a fine processing module and a control module; the conveying module comprises a conveyor belt (1); the fixing module comprises a positioning platform (3) fixed to the conveyor (1), and an agarwood clamp (7) fixed to the positioning platform (3) fixes a single piece of wood (8); the nuclear magnetic resonance scanning device (4) surrounds the conveyor belt (1); the image processing and control module (2) is located on the side of the conveyor belt; the rough processing module comprises a rough processing tool device (5) fixed inside the conveyor belt (1); the fine processing module comprises a multi-axis engraving machine (6) fixed to the side of the conveyor belt.

2. The automated agarwood separation and processing system based on nuclear magnetic resonance imaging according to claim 1, characterized in that: The conveyor belt (1) has multiple independently adjustable speeds and is made entirely of non-metallic materials and non-ferromagnetic metal materials. The speed of the conveyor belt (1) is matched with the scanning speed of the magnetic resonance imaging device (4) through the magnetic resonance imaging device (4), and the speed of the conveyor belt (1) is adjusted according to the cutting speed through the rough machining tool device (5).

3. The automated agarwood separation and processing system based on nuclear magnetic resonance imaging according to claim 1, characterized in that: The positioning platform (3) and the agarwood clamp (7) are made of non-metallic materials and non-ferromagnetic metal materials. The positioning platform (3) moves vertically and rotates along the conveyor belt. The agarwood clamp (7) is composed of a plurality of groups of clamps and the material used has a different imaging effect than wood in the nuclear magnetic resonance scanning device (4).

4. The automated agarwood separation and processing system based on nuclear magnetic resonance imaging according to claim 1, characterized in that: During the processing stage of the multi-axis engraving machine (6), the wood (8) clamped by the positioning table (3) is rotated to a direction perpendicular to the ground by rotating the fixed table.

5. The automated agarwood separation and processing system based on nuclear magnetic resonance imaging according to claim 1, characterized in that: The movement paths of the rough machining tool equipment (5), the multi-axis engraving machine (6) and the positioning platform (3) are all automatically generated by the image processing and control module (2).

6. A method for automated separation and processing of agarwood based on nuclear magnetic resonance imaging, characterized in that: The automated agarwood separation and processing system based on nuclear magnetic resonance imaging as described in any one of claims 1 to 5 comprises the following steps: S1: Use magnetic resonance imaging to obtain high-resolution three-dimensional images of agarwood. S2: Input the acquired high-resolution three-dimensional stereo image into the image processing and control module for the first fitting, and preliminarily fit the three-dimensional image into multiple planes perpendicular to the running direction of the conveyor belt. S3: The second fitting is a more fine-grained processing. Based on the first step of fitting, it is further fitted into a polyhedron composed of multiple smaller planes to ensure high-precision surface matching. S4: Based on the polyhedron model after two fittings, the combined motion path of the rough processing tool equipment (5) and the fine processing multi-axis engraving machine (6) is automatically generated to achieve efficient and accurate processing of agarwood. S5: Execute the above generated path, perform preliminary shape rough processing through rough processing tool equipment, and then use a multi-axis engraving machine to perform detailed fine processing to complete the final processing of the agarwood.

7. The method for automated separation and processing of agarwood based on nuclear magnetic resonance imaging according to claim 6, characterized in that: The first step of fitting the high-resolution agarwood three-dimensional stereo image in S2 includes: first, segmenting the internal agarwood at a fixed equidistant, then using the maximum intensity projection technology for the three-dimensional data of each small segment of the agarwood log, projecting the three-dimensional image data onto the tangential plane of the agarwood log, and finding the largest cross-section; then calculating the minimum circumscribed rectangle enclosing multiple combinations of maximum cross-sections in three-dimensional space, and using the minimum circumscribed rectangle as the cutting surface.

8. The method for automated separation and processing of agarwood based on nuclear magnetic resonance imaging according to claim 6, characterized in that: The second fitting in S3 performs finer-grained processing and includes the following steps: first, denoising the high-resolution three-dimensional image; then, using a three-dimensional convex hull algorithm, generating an initial polyhedron surrounding the agarwood; then, using a mesh deformation algorithm combined with a gradient descent method, iteratively optimizing the vertex positions of the polyhedron so that it more accurately fits the surface shape of the agarwood.

9. The method for automated separation and processing of agarwood based on nuclear magnetic resonance imaging according to claim 6, characterized in that: The S4 is based on the polyhedron model after two fittings. The polyhedron model after the second fitting is inside the polyhedron model after the first fitting, and the outer surface of the log is based on the polyhedron after the first fitting.

10. The method for automated separation and processing of agarwood based on nuclear magnetic resonance imaging according to claim 6, characterized in that: After the S4 automatically generates the combined motion path of the rough machining tool device (5) and the fine machining multi-axis engraving machine (6), the processing time of the rough machining tool device (5) and the fine machining multi-axis engraving machine (6) is automatically calculated, and the image processing and control module (2) automatically adjusts the scanning speed and resolution of the nuclear magnetic resonance scanning device (4).

Images