Cooperative spinning behavior trajectory data acquisition method for silkworms

Abstract

The invention discloses a cooperative spinning behavior trajectory data acquisition method for silkworms, which comprises the following steps: firstly, identifying the heads of the silkworms; secondly, dynamic imaging is conducted on the spinning process of silkworms in the silkworm cocoon forming process through a CT scanning system, and a behavior image set is obtained; then, positioning the spatial position of the silkworm head in the image based on a preset image processing algorithm; and finally, through a multi-dimensional data fusion technology, constructing a three-dimensional dynamic trajectory model of the cooperative spinning behavior of the silkworms, and generating a trajectory video and a behavior trajectory diagram. According to the invention, nondestructive testing of cooperative spinning behaviors of one or more silkworms in the silkworm cocoons is realized; cT is combined with a deep learning algorithm, so that the positioning precision is greatly improved; and multi-scale behavior analysis from a single silkworm to a group is realized. According to the method, the labor intensity of manual observation is remarkably reduced. According to the method, a three-dimensional dynamic imaging technology is introduced, positioning and continuous tracking of single or multiple silkworm spinning behavior tracks in silkworm cocoons are achieved, and a brand new technical means is provided for silkworm spinning behavior research.

Description

Technical Field

[0001] The present invention relates to the field of data acquisition, and more specifically, to a method for acquiring data on the collaborative silk-spinning behavior trajectories of silkworms. Background Art

[0002] Silkworm cocoons, as a unique natural composite material, have attracted much attention due to their excellent physical properties and biocompatibility. However, due to their small size, irregular shape, and complex internal structure, single silkworm cocoons are difficult to directly meet the requirements of complex engineering structures for material properties. In recent years, studies have found that composite cocoons formed by multiple silkworms collaborating in silk-spinning not only exhibit a more complex and unique internal structure, such as multi-layer winding patterns, fiber arrangement rules, and a significant improvement in mechanical properties, but also provide new possibilities for the engineering application of silkworm cocoon materials. This discovery indicates that in-depth research on the formation mechanism and behavioral characteristics of the collaborative silk-spinning behavior of multiple silkworms, especially the analysis of the interaction relationships between environmental factors and the cocoon-spinning behavior of silkworms, genetic characteristics and behavioral patterns, is of great theoretical and practical significance for optimizing the quality of silkworm cocoon materials, enhancing their functionality, and designability.

[0003] Obtaining the silk-spinning behavior trajectories of silkworms is crucial for understanding the process of silk fibroin formation, and it also has important value and practical significance for analyzing the group behavior rules and formation mechanisms of multiple silkworms collaborating in silk-spinning to form cocoons. However, current research on the collaborative silk-spinning behavior of silkworms still faces many challenges. In the initial stage of silk-spinning, the behavioral dynamics of silkworms can be photographed and observed with the naked eye or multiple cameras. However, once the silkworm cocoon is formed, its internal structure becomes opaque and complex, and traditional observation methods (such as the naked eye or a camera) cannot capture the dynamic behavior of silkworms inside the cocoon, resulting in difficulties in subsequent data collection of behavioral trajectories. Current technical means can only collect silk-spinning behavior data before the formation of silkworm cocoons, and the accuracy and spatio-temporal resolution of the data are relatively low, making it difficult to comprehensively reflect the dynamic laws of silkworm silk-spinning behavior. Once the silkworm cocoon is formed, there is still no effective solution for subsequent behavioral trajectories, which severely limits the complete recording and in-depth research of the collaborative silk-spinning behavior of multiple silkworms.

[0004] In the prior art, although some studies have attempted to use optical imaging or other non-invasive detection methods to observe and record the silk-spinning behavior of silkworms, these methods still have obvious limitations in the depth and accuracy of data acquisition due to the opacity and complexity of the internal structure of the cocoon. For example, the invention patent with publication number CN103808308A discloses a method for automatically collecting silk-spinning behavior data of silkworms, which uses a collection of behavioral images of silkworms spinning on a camera acquisition platform. However, optical imaging technology is difficult to penetrate the cocoon, and it is impossible to obtain complete information on the dynamic behavior of silkworms spinning inside the cocoon after the cocoon is formed; and although traditional tomography technology can provide images with a certain spatial resolution, its time resolution and dynamic tracking capabilities are limited, and it is difficult to meet the needs of continuous and accurate recording of the coordinated silk-spinning behavior of multiple silkworms. In addition, the silkworm has a large amount of movement during the silk-spinning process, and the morphological characteristics of its head and tail make it difficult for simple CT scanning to accurately identify and track the behavior trajectory, further exacerbating the difficulty of data collection.

[0005] Therefore, there is an urgent need for a technical solution that can break through the limitations of traditional methods and achieve non-destructive, high-precision dynamic tracking and analysis of the coordinated silk-spinning behavior of multiple silkworms inside cocoons.

[0006] In summary, in response to the shortcomings of the prior art, the present invention proposes a method for collecting trajectory data of the collaborative silk-spinning behavior of silkworms based on high-resolution CT scanning and intelligent image processing, aiming to reveal the dynamic laws of the collaborative silk-spinning behavior of multiple silkworms and its influence mechanism on the performance of silkworm cocoon materials through non-destructive testing and high-precision positioning technology, so as to provide a scientific basis and technical support for the optimal design and engineering application of silkworm cocoon materials. Summary of the invention

[0007] In view of this, the present invention provides a method for collecting trajectory data of silkworm cooperative silk-spinning behavior, which is used to solve the problem that accuracy and consistency are difficult to ensure in existing methods for collecting trajectory data of silkworm cooperative silk-spinning behavior.

[0008] To achieve the above objectives, the proposed solution is as follows: A method for collecting trajectory data of silkworm cooperative silk-spinning behavior, comprising: Marking the heads of each silkworm in the cocoon sample; The cocoon samples were fixed and continuously CT scanned to obtain a collection of behavioral images; Based on a preset image processing algorithm, the spatial position of each silkworm head in each image frame in the behavioral image set is identified; Through multi-dimensional data fusion technology, a three-dimensional dynamic trajectory model of the silkworm's coordinated silk-spinning behavior was constructed based on the time and spatial positions of the marking points, and the trajectory map and trajectory video of each silkworm's silk-spinning behavior were generated.

[0009] Preferably, after drawing based on the temporal and spatial positions of the marker points, the following steps are further included: The completed trajectory and the trajectory video are stored in real time.

[0010] Preferably, the process of identifying the spatial positions of the silkworm heads in each image frame of the behavior image set based on a preset image processing algorithm includes: Preprocess the image frames, including grayscale adjustment, contrast enhancement, and denoising; Perform end-to-end processing on each image frame in the behavior image set through the YOLOv8 object detection algorithm, and identify the bounding boxes of the silkworm heads in each image; YOLOv8 uses a convolutional neural network to extract features from the input image, and performs object classification and localization through multiple convolutional layers and fully connected layers; Extract the bounding boxes of each silkworm head from the YOLOv8 detection results, and combine their spatial coordinates with the time series to determine the spatial positions of each silkworm head; Initialize the detected spatial positions of the silkworm heads, and assign a unique identifier to each silkworm; Use the Kalman filter to predict the movement trajectory of the silkworm heads, and predict the next position of the silkworms according to the movement state of the targets; Match the targets in the front and rear frames through the Hungarian algorithm; Update the spatial positions of the silkworm heads in real time according to the detected new positions and the prediction results.

[0011] Preferably, the process of fixing the silkworm cocoon sample and continuously performing CT scanning includes: Place the silkworm cocoon sample inside the fixing device in the micro-CT scanning stage for fixing; The micro-CT scanning stage emits X-rays to penetrate the fixed silkworm cocoon sample, and performs CT scanning on the silkworm cocoon sample according to the preset voltage and current conditions, frame rate, and resolution.

[0012] Preferably, before identifying the spatial positions of the silkworm heads in each image frame of the behavior image set based on a preset image processing algorithm, the following steps are further included: Perform grayscale adjustment, contrast enhancement, and format conversion on each image frame in the behavior image set.

[0013] According to the specific embodiments provided by the present invention, the following technical effects are disclosed by the present invention: The method for collecting data on the collaborative silk-spinning behavior trajectories of silkworms provided by the present invention innovatively introduces three-dimensional dynamic imaging technology and for the first time realizes the precise positioning and continuous tracking of the silk-spinning behavior trajectories of single or multiple silkworms inside the cocoon. First, the heads of the silkworms in the cocoon samples are marked; then, the cocoon samples are fixed, and high-resolution CT scanning systems are used to continuously perform CT scans to obtain a set of silk-spinning behavior images of the silkworms; based on a preset image processing algorithm, the spatial positions of the heads of the silkworms in each image frame of the set of behavior images are identified; finally, through multi-dimensional data fusion technology, a three-dimensional dynamic trajectory model of the collaborative silk-spinning behavior of the silkworms is constructed based on the temporal and spatial positions of the marked points, and trajectory diagrams and trajectory videos of the silk-spinning behaviors of the silkworms are generated. The trajectory data collection method of the present invention uses high-resolution CT combined with deep learning algorithms for non-destructive detection of the collaborative silk-spinning behaviors of single or multiple silkworms inside the cocoon, realizes multi-scale behavior analysis from single silkworms to groups, reduces the manual labor burden, and improves the efficiency of collecting data on the silk-spinning behaviors of silkworms. At the same time, the accuracy and consistency of data collection are improved. Moreover, it provides a brand-new technical means and method strategy for the research on the silk-spinning behaviors of silkworms, and has important theoretical value and practical application prospects. Brief Description of the Drawings

[0014] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only the embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained according to the provided drawings without creative efforts.

[0015] Figure 1 It is a flowchart of a method for collecting data on the collaborative silk-spinning behavior trajectories of silkworms provided by an embodiment of the present invention; Figure 2 It is a schematic diagram of another method for collecting data on the collaborative silk-spinning behavior trajectories of silkworms provided by an embodiment of the present invention; Figure 3 It is a schematic diagram of a set of silk-spinning behavior images of silkworms inside a cocoon provided by an embodiment of the present invention. Detailed Description of the Embodiments

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0017] First, in combination with Figure 1-2An introduction is given to a method for collecting data on the collaborative silk-spinning behavior trajectories of silkworms provided by an embodiment of the present invention, as Figure 1-2 shown. The method includes: Step S01, identifying the heads of each silkworm in the silkworm cocoon sample.

[0018] Specifically, first identify the heads of the silkworms. After the silkworms enter the cocoon-spinning state, collect the collaborative silk-spinning behavior trajectories of the silkworms. For example: Based on the difference in material density, small steel balls can be pasted on the heads of the silkworms as markers. Other marking methods can also be used to improve the recognizability and tracking effect of the silkworm heads in subsequent images.

[0019] Step S02, fix the silkworm cocoon sample and continuously perform CT scans.

[0020] Specifically, use a high-resolution CT scanning system (high-precision microfocus X-ray three-dimensional scanning system) to dynamically image the silk-spinning process of the silkworms during the formation of the silkworm cocoons. Based on the micro-CT scanning technology, photograph the process of the silkworms forming cocoons to obtain the original data of the silkworms during the cocoon-forming period. Place the silkworms in the cocoon-forming process on the micro-CT scanning table, keep the silkworm cocoon sample fixed, and continuously photograph the cocoon-forming process to obtain a set of behavior images. As Figure 3 shown, perform CT scans on five silkworm cocoons to obtain a set of silk-spinning behavior images of the silkworms inside the five silkworm cocoons.

[0021] Place the silkworm cocoon sample inside the fixing device in the micro-CT scanning table for fixation. A silkworm cocoon fixing device using foam material as a support can be used. The fixing device can effectively fix the silkworm cocoon within the scanning area of the micro-CT scanning table (microCT scanner), ensuring the stability of the silkworm cocoon during the scanning process and avoiding the impact of movement or deformation on the imaging quality.

[0022] The micro-CT scanning table emits X-rays to penetrate the fixed silkworm cocoon sample and continuously performs CT scans on the silkworm cocoon sample at a preset frame rate and resolution. The frame rate and resolution of the CT scan can be dynamically adjusted according to the characteristics of the silkworm silk-spinning behavior to ensure that the recorded silkworm movement trajectories are complete and continuous and adapt to the silk-spinning speeds and behavior changes at different stages. In addition, the micro-CT scanning table has rotation and lifting functions and can perform multi-angle scans by adjusting the table angle to ensure obtaining images of the silkworm cocoon-forming process from different perspectives. The micro-CT scanning table can perform 360° omnidirectional imaging of the internal structure of the silkworm cocoon sample, thereby generating complete and accurate data on the silk-spinning behavior trajectories of the silkworms.

[0023] Step S03, identify the spatial positions of the heads of each silkworm in each image frame in the set of behavior images based on a preset image processing algorithm.

[0024] Specifically, the sub-pixel level positioning of the head spatial positions of each silkworm in the set of behavior images can be based on deep learning algorithms, achieving a positioning accuracy of millimeters. For example, the YOLOv8 object detection algorithm and the DeepSort object tracking algorithm can be preset to detect and track the movement of the key points of the silkworm head.

[0025] First, the image frames in the image set can be preprocessed, including grayscale adjustment, contrast enhancement, and denoising, etc.

[0026] Then, the YOLOv8 object detection algorithm can be used to detect the identifiers of each silkworm head in each image frame in the set of behavior images, and the spatial position of the silkworm head can be captured in real time. The YOLOv8-pose model can be used to perform object detection on multiple silkworms, and the head marker points of each silkworm can be obtained to capture the posture changes and movement trajectories when the silkworm spins silk. Through the key point connection information, the silk-spinning behavior process of each silkworm can be obtained, and the collaborative process and silk-spinning rules can be analyzed. The specific process is as follows: The YOLOv8 object detection algorithm is used to perform end-to-end processing on each frame of the image, the bounding box of the silkworm head is identified in the image, and the spatial position of each silkworm is obtained through the neural network. YOLOv8 uses a convolutional neural network (CNN) to extract features from the input image, and performs object classification and positioning through multiple convolutional layers and fully connected layers; the identification box of each silkworm head is extracted from the YOLOv8 detection results, and its spatial coordinates are combined with the time series to determine the spatial position of each silkworm head.

[0027] The DeepSort object tracking algorithm performs trajectory prediction and update based on the spatial position of the silkworm head detected by the YOLOv8 object detection algorithm, realizes the tracking of the movement trajectory of the silkworm head, and obtains the real-time spatial position of the silkworm head. Using the DeepSort object tracking algorithm to perform object tracking on multiple silkworms can solve problems such as occlusion and overlap of multiple silkworms at the same time. The DeepSort (Deep Sorting) algorithm combines deep learning feature extraction and Kalman filter prediction, etc. Each silkworm is continuously identified and tracked through appearance features, Kalman filter, and Hungarian algorithm. DeepSort can assign a unique identifier to each silkworm in real time to ensure continuous tracking of the behavior trajectory throughout the observation process. The specific process is as follows: The Kalman filter is used to predict the movement trajectory of the silkworm head, and the next position of the silkworm is predicted according to the movement state of the target (such as speed and acceleration) to ensure that the target can still be tracked in the case of fast movement or occlusion. The Hungarian algorithm is used for object matching to solve the object association problem when multiple objects are occluded. The objects in the front and back frames are matched through the algorithm to ensure the uniqueness and continuity of the object. According to the detected new position and the prediction result, the spatial position of the silkworm head is updated in real time to ensure the continuity and accuracy of the movement trajectory.

[0028] The head of the silkworm in the image is accurately located through the YOLOv8 object detection algorithm. At the same time, the DeepSort object tracking algorithm is combined to continuously track multiple silkworms. The combination of the two can achieve the precise detection of multiple silkworm individuals, key point calibration, and efficient tracking of the behavior trajectories, revealing the movement laws of multiple silkworms during the collaborative silk-spinning process.

[0029] Step S04: Through the multi-dimensional data fusion technology, a three-dimensional dynamic trajectory model of the collaborative silk-spinning behavior of silkworms is constructed based on the temporal and spatial positions of the marked points, generating the trajectory maps and trajectory videos of the silk-spinning behaviors of each silkworm.

[0030] Specifically, according to the silkworm head marked points detected by the YOLOv8 object detection algorithm and the tracking of the silkworm head marked points by the DeepSort object tracking algorithm, the trajectory maps and trajectory videos of each silkworm can be obtained by connecting the spatial positions of the marked points according to the time series corresponding to the image frames.

[0031] The acquisition method provided by the embodiments of the present invention realizes the instant acquisition of data and supports the intuitive display of the acquisition results in the form of pictures and videos.

[0032] Through the movement trajectories of multiple silkworm marked points, the collaborative silk-spinning behaviors of multiple silkworms can be analyzed. Trajectory clustering and statistical analysis methods can be used to mine the behavior patterns of multiple silkworms during the collaborative silk-spinning process. For example: moving in the same direction and aggregating, dispersing movement, or aggregation progress, etc. In addition, the distribution of the silk-spinning paths of silkworms, the changes in relative positions, and the time periodicity and other characteristics can also be analyzed according to the trajectories of multiple silkworms, revealing their behavior laws. For example: whether there are cooperation, mutual interference, or queuing patterns when silkworms spin silk. Combining the speed, direction, and path complexity of the trajectories can further reveal the behavior patterns of multiple silkworms during the collaborative silk-spinning process, such as collective movement, synchronous movement, etc.

[0033] The embodiments of the present invention can accurately track the silk-spinning behaviors of silkworms through the combination of the YOLOv8 object detection algorithm and the DeepSort object tracking algorithm. Especially in the case of multiple silkworms spinning silk collaboratively, it can obtain the collaborative movement trajectories of multiple silkworms when jointly constructing a cocoon and analyze the behavior patterns of the silkworm group during collaborative silk-spinning, revealing the cooperative relationships and silk-spinning strategies among silkworms.

[0034] The method for collecting data on the collaborative silk-spinning behavior trajectories of silkworms provided by the embodiments of the present invention innovatively introduces three-dimensional dynamic imaging technology and for the first time realizes the precise positioning and continuous tracking of the silk-spinning behavior trajectories of single or multiple silkworms inside the cocoon. First, the heads of the silkworms in the cocoon sample are marked; then, the cocoon sample is fixed, and a high-resolution CT scanning system is used to continuously perform CT scans to obtain a set of silk-spinning behavior images of the silkworms; based on a preset image processing algorithm, the spatial positions of the heads of the silkworms in each image frame of the set of behavior images are identified; finally, through multi-dimensional data fusion technology, a three-dimensional dynamic trajectory model of the collaborative silk-spinning behavior of the silkworms is constructed based on the temporal and spatial positions of the marked points, and trajectory maps and trajectory videos of the silk-spinning behaviors of the silkworms are generated. The trajectory data collection method of the embodiments of the present invention uses high-resolution CT combined with deep learning algorithms for non-destructive detection of the collaborative silk-spinning behaviors of single or multiple silkworms inside the cocoon, realizes multi-scale behavior analysis from single silkworms to groups, reduces the manual labor burden, and improves the efficiency of collecting data on the silk-spinning behaviors of silkworms. At the same time, the accuracy and consistency of data collection are improved. Moreover, it provides a brand-new technical means and method strategy for the study of the silk-spinning behaviors of silkworms, and has important theoretical value and practical application prospects.

[0035] In order to better perform behavior analysis and production process optimization based on the behavior trajectory data, after obtaining the silkworm trajectories by plotting based on the temporal and spatial positions of the marked points in step S04 of the embodiments of the present invention, the following steps may further be executed: The completed trajectory maps and trajectory videos are stored in real time.

[0036] Specifically, a data storage module is set up to save and manage the motion trajectory data of the silk-spinning behaviors of silkworms, and the silk-spinning behavior data of the heads of silkworms during the silk-spinning process is saved in real time. At the same time, multiple data export options may be provided.

[0037] Further, in order to be able to perform more precise identification, on the basis of the method for collecting data on the collaborative silk-spinning behavior trajectories of silkworms provided by the foregoing embodiments, before identifying the spatial positions of the heads of the silkworms in each image frame of the set of behavior images based on a preset image processing algorithm in step S03, the following steps may further be executed: The gray levels of each image frame in the set of behavior images are adjusted, the contrast is enhanced, and the format is converted.

[0038] Finally, it should also be noted that in this text, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, article or device comprising the element.

[0039] The various embodiments in this specification are described in a progressive manner. Each embodiment focuses on the differences from other embodiments. For the same or similar parts among the various embodiments, reference can be made to each other.

[0040] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the broadest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for collecting data on the collaborative silk-spinning behavior trajectories of silkworms, characterized in that, Comprising: Identifying the heads of each silkworm in the cocoon sample; Fixing the cocoon sample and continuously performing CT scans to obtain a set of behavior images; Identifying the spatial positions of the heads of each silkworm in each image frame in the set of behavior images based on a preset image processing algorithm; Through multi-dimensional data fusion technology, constructing a three-dimensional dynamic trajectory model of the collaborative silk-spinning behavior of silkworms based on the temporal and spatial positions of the marked points, and generating trajectory maps and trajectory videos of the silk-spinning behavior of each silkworm.

2. The method for collecting data on the collaborative silk-spinning behavior trajectory of silkworms according to claim 1, wherein, After drawing based on the temporal and spatial positions of the marked points, it further includes: Storing the completed trajectory and trajectory video in real time.

3. The method for collecting data on the collaborative silk-spinning behavior trajectory of silkworms according to claim 2, wherein The identifying the spatial positions of the heads of silkworms in each image frame in the set of behavior images based on a preset image processing algorithm includes: Preprocessing the image frames, including grayscale adjustment, contrast enhancement, and noise reduction; Performing end-to-end processing on each image frame in the set of behavior images through the YOLOv8 object detection algorithm, and identifying the identification frames of the heads of silkworms in each image; YOLOv8 uses a convolutional neural network to extract features from the input image, and performs object classification and localization through multiple convolutional layers and fully connected layers; Extracting the identification frames of each silkworm head from the YOLOv8 detection results, and combining its spatial coordinates with the time series to determine the spatial position of each silkworm head; Initializing the detected spatial positions of the silkworm heads, and assigning a unique identifier to each silkworm; Using Kalman filtering to predict the movement trajectory of the silkworm heads, and predicting the next position of the silkworms according to the movement state of the targets; Matching the targets in the front and back frames through the Hungarian algorithm; According to the detected new positions and prediction results, updating the spatial positions of the silkworm heads in real time.

4. The method for collecting data on the collaborative silk-spinning behavior trajectory of silkworms according to claim 1, characterized in that, The process of fixing the cocoon sample and continuously performing CT scans includes: Placing the cocoon sample inside the fixing device in the micro-CT scanning table for fixing; The micro-CT scanning table emits X-rays to penetrate the fixed cocoon sample, and performs CT scans on the cocoon sample according to the preset voltage and current conditions, frame rate, and resolution.

5. The method for collecting data on the collaborative silk-spinning behavior trajectory of silkworms according to any one of claims 1 to 4, characterized in that, Before identifying the spatial positions of the heads of each silkworm in each image frame in the set of behavior images based on a preset image processing algorithm, it further includes: Performing grayscale adjustment, contrast enhancement, and format conversion on each image frame in the set of behavior images.

Images