CN-121999488-A - Device based on disease medium biological database and application method

CN121999488ACN 121999488 ACN121999488 ACN 121999488ACN-121999488-A

Abstract

The invention provides equipment and an application method based on a disease medium biological database, wherein the equipment comprises a fixed component and an acquisition terminal, the fixed component comprises a storage table, a control bracket is arranged on the storage table, the acquisition terminal is arranged on the control bracket, the acquisition terminal is provided with an identification module, a marking module, a counting module and a classification module, a sample is firstly placed on the storage table, then the position of the acquisition terminal is controlled through the control bracket so that the acquisition terminal samples the sample, and finally information acquired by sampling is respectively identified through the identification module, marked by the marking module and classified by the classification module. The invention can collect, acquire, intelligently identify, mark, accurately count and classify samples, and realize full-process automation and intellectualization, thereby shortening the working time for screening the disease medium organisms from the large-flux samples, and further greatly improving the timeliness, accuracy and intellectualization level of public health monitoring, port import and export goods clearance and disease medium organism prevention and control.

Inventors

LI TINGTING
CHEN JIAN
QIU DEYI
LIU DEXING
WEI XIAOYA
WU SIWEI

Assignees

中山海关技术中心

Dates

Publication Date: 20260508
Application Date: 20260122

Claims (10)

1. An apparatus based on a disease vector biological database, comprising: the fixing assembly comprises a storage table, and a control bracket is arranged on the storage table; the acquisition terminal is arranged on the control bracket and is provided with an identification module, a marking module, a counting module and a classification module; The sample is firstly placed on the object placing table, then the position of the acquisition terminal is controlled through the control bracket so that the acquisition terminal samples the sample, and finally the information obtained by sampling is respectively identified by the identification module, marked by the marking module and classified by the classification module.
2. The disease media biological database-based device of claim 1, wherein the control stent comprises a clamping stent and a fixed stent; the clamping support comprises a telescopic mechanical arm and a clamping member, the clamping member is arranged at one end of the telescopic mechanical arm, the clamping member consists of two parts, one part is provided with a gear bolt, the other part is provided with a saw tooth slot, and the gear bolt is matched with the saw tooth slot; The telescopic support is characterized in that a telescopic support rod is arranged in the fixed support, one end of the telescopic support rod is connected in the fixed support through a spherical structural part, a threaded fixing piece is further arranged in the fixed support, lines for adapting to the threaded fixing piece are arranged on the surface of the telescopic support rod, a screw knob is further arranged on the threaded fixing piece, the screw knob is rotated to control tightness of the telescopic support rod and the threaded fixing piece, and a plastic gasket is arranged at the other end of the telescopic support rod.
3. The device based on a disease medium biological database according to claim 1, wherein a sample tray is arranged at the bottom of the object placing table, a pull ring is arranged at one end of the sample tray, two sides of the other end of the sample tray incline upwards, and an opening is formed in the middle of the other end of the sample tray.
4. The device based on a disease medium biological database according to claim 1, wherein the inner side wall of the object placing table is provided with a surrounding type variable temperature lamp, the surrounding type variable temperature lamp comprises an inner lamp belt and an outer soft light cover, and the outer soft light cover is made of semitransparent frosted materials.
5. The device based on a disease medium biological database according to claim 1, wherein the inner wall of the bracket connected with the control bracket of the object placing table is provided with a groove, an embedded lamp is arranged in the groove, the embedded lamp comprises an inner lamp source and an outer lamp shade, the embedded lamp can be ejected or fixed by pressing, and the embedded lamp is provided with a foldable and rotatable bracket.
6. The apparatus based on a disease media biological database according to claim 2, wherein the gear pins are cross-shaped, the tip circle thickness of the gear pins is 0.1mm, and the interval between each serration of the serration grooves is 0.15mm.
7. The device based on a disease media biological database according to claim 1, wherein the acquisition terminal is a mobile phone with a camera function, and an application program comprising the identification module, the marking module, the counting module and the classification module is installed in the mobile phone.
8. A method of using a device based on a disease vector biological database, characterized in that it is carried out with a device according to any one of claims 1 to 7, comprising the specific steps of: creating a task queue; Sample picture information is obtained, identified through a preset algorithm service layer and put into a task queue; preprocessing sample picture information in a task queue; constructing a deep learning model, connecting the deep learning model with a disease medium biological database and training; analyzing and comparing the preprocessed sample picture information through a trained deep learning model; and generating a report from the analysis and comparison results, and marking key information in the report.
9. The method for applying a device based on a disease media biological database according to claim 8, wherein the preprocessing of the sample picture information in the task queue further comprises the following specific steps: constructing YoLo n model; Loading YoLon.pt weight to finely adjust YoLo n model; testing the finely-adjusted YoLo n model by matching with the built deep learning detection framework; Importing the tested YoLo n model into the preprocessed sample picture information for counting; And generating a report of the counting result.
10. The method for applying a device based on a disease media biological database according to claim 8, wherein the preprocessing of the sample picture information in the task queue further comprises the following specific steps: Constructing a multi-category image classification model; the multi-category image classification model is trained on an ImageNet data set; Leading the pre-trained model into the pre-processed sample picture information for classification; verifying the classification result through a cross entropy loss function and AdamW optimization algorithm; and generating a report according to the verified classification result.

Description

Device based on disease medium biological database and application method Technical Field The invention relates to the technical field of disease medium biology, in particular to a device based on a disease medium biological database and an application method thereof. Background Disease-causing organisms such as mosquitoes, flies, mice, ticks, cockroaches, etc., are key agents for the transmission of a variety of major infectious diseases such as dengue fever, malaria, plague, lyme disease, etc. The method is used for continuous, accurate and efficient monitoring and population dynamic analysis, and is a foundation for epidemic early warning and scientific prevention and control in modern public health systems. The traditional disease medium biological monitoring workflow highly depends on manual work, and the main mode is that a trapping device (such as a mosquito trapping lamp and a mouse sticking plate) is arranged on site, samples are periodically recovered, manual identification, classification and counting are carried out by professionals under a microscope in a laboratory according to morphological characteristics, and data are manually recorded and summarized. This traditional model, when dealing with increasingly complex public health challenges, exposes the following inherent bottlenecks and challenges: 1. strong professional dependence and low efficiency, and specific technicians who need to be trained for a long time for accurately identifying specific species. The manual identification process is time-consuming and labor-consuming, the sample processing flux is low, and human resources become key constraints when facing large-scale monitoring or epidemic emergency response, so that the monitoring period is long, and the data output is slow. 2. The subjective error and consistency problems are that the identification result is easily influenced by subjective factors such as personnel experience, fatigue degree and the like, and the judgment of different personnel and even the same person on the same batch of samples at different times can be different, so that the consistency and objectivity of data are difficult to ensure, and the reliability of long-term trend analysis is influenced. 4. The counting statistics is tedious and error-prone, for high-density samples (such as a large number of mosquitoes collected by a mosquito-lured lamp), the manual counting is huge in workload, and the counting error is easily caused by visual fatigue or individual adhesion, so that the accuracy of the key index of population density is affected. In recent years, with the rapid development of computer vision and artificial intelligence technology, especially the breakthrough of deep learning in the field of general object recognition and classification, a new technical path is provided for solving the problems. Some common image recognition applications have appeared on the market, as well as agricultural and industrial oriented automated counting devices. However, there are significant limitations to applying these techniques directly to the field of medical biomonitoring professions: 1. The field specificity is lacking, the general recognition model lacks a special database optimized for the fine morphological characteristics (such as veins, body surface bristles, mouth parts structures and the like) of the vector organism, and is difficult to realize high-precision category-category classification, and the severe requirements of infectious disease tracing on the precise identification of vector species cannot be met. 2. The scene suitability is insufficient, namely the precision of the general target detection and counting algorithm is rapidly reduced under the disordered background, and the missed detection, false detection and segmentation errors are easy to occur. 3. The service closed loop is not opened, most of the prior art is a software tool or algorithm module with single function, and the prior art cannot be deeply fused with a disease medium monitoring and standardized service flow and public health data management system to form an intelligent solution of 'end-to-end'. The identification result cannot be automatically bound with the space-time information, and is difficult to directly interface with the existing disease medium biological monitoring information management system. Disclosure of Invention The invention aims to identify, mark, count and classify collected samples based on a disease medium biological database, and the specific implementation scheme comprises 1, solidifying a shooting background, reducing missed detection, false detection and segmentation errors caused by rapid decline of precision in operation, 2, deeply fusing disease medium biological expertise, an artificial intelligent core algorithm and an Internet of things hardware technology based on the disease medium biological database, and utilizing the specialized disease medium biological image characteristic database as a