White tablet recognition system and method based on short-wave infrared focal plane detector

Through a system based on short-wave infrared focal plane detector, using short-wave infrared light and deep neural network to identify white pills, the problem of difficulty in distinguishing the types of white pills in traditional detection methods is solved, and efficient and accurate pill recognition is achieved.

CN118937266APending Publication Date: 2024-11-12SUZHOU JIEZHUN INTELLIGENT TECH R & D CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411043472.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-31
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

Traditional tablet detection methods cannot accurately detect the types of tablets due to the influence of the external environment, especially white tablets with similar colors and shapes, which lead to misjudgment and confusion.

Method used

The system based on the short-wave infrared focal plane detector is adopted to emit short-wave infrared light of 1450nm-1650nm wavelength through the infrared light source, and the signal is focused and processed by using the focus module and the short-wave infrared detector, and combined with the deep neural network to train and identify the image data. Use grayscale histograms to distinguish different types of pills.

Benefits of technology

It realizes rapid and accurate detection of white pills, improves the sensitivity and accuracy of detection, can distinguish slight differences between different batches, reduces the misjudgment rate, and enhances the stability and reliability of detection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118937266A_ABST
    Figure CN118937266A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of optical detection, in particular to a white tablet recognition system based on a short-wave infrared focal plane detector, which comprises an infrared light source, a short-wave infrared detector, a focusing module, a main control module, a data processing module and a display module, the white tablet identification method based on the short-wave infrared focal plane detector comprises the following steps: S1, emitting infrared light; s2, infrared light receiving: the infrared light reflected in the step S1 passes through a focusing module, and the infrared light is focused on a short wave infrared detector; s3, analog signal generation: the main control module drives the short-wave infrared detector to generate pixel current, and the pixel current is converted into an analog signal through a circuit in the short-wave infrared detector; s4, data acquisition and analog-to-digital conversion: receiving an analog signal and converting the analog signal into a digital signal; and S5, data processing and image display: receiving the digital signal converted in the step S4 by the main control module, carrying out algorithm processing on the digital signal, and then displaying an image in a display module.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of optical detection technology, and in particular to a white pill recognition system and method based on a short-wave infrared focal plane detector. Background Art

[0002] Traditional tablet detection methods usually rely on visual inspection and manual classification, which are subjective and unstable. Other optical methods such as ordinary infrared imaging technology are greatly affected by ambient light and are not suitable for accurate detection in complex scenarios.

[0003] Many tablets may be very similar in appearance, especially white tablets with similar colors and shapes. Traditional visual inspection methods have difficulty distinguishing them, which can easily lead to misjudgment or confusion. There may be batch differences or minor manufacturing changes in the drug production process, which may affect the appearance of the tablets, making the surface characteristics of the same type of tablets slightly different between batches, increasing the difficulty of detection.

[0004] Short-wave infrared light is a type of electromagnetic wave with a wavelength between 760nm and 14um. It is outside the visible light spectrum and cannot be recognized by the naked eye, but it can be detected and imaged by specific sensors (such as InGaAs). Short-wave infrared light has some special properties and applications. It can interact with objects in a similar way to visible light to produce shadows and contrasts, thereby providing an effective method for object recognition.

[0005] Therefore, the present application has developed a white pill recognition system and method based on a short-wave infrared focal plane detector to solve the problems existing in the prior art. Summary of the invention

[0006] The purpose of the present invention is to provide a white pill recognition system and method based on a short-wave infrared focal plane detector to solve the problem that the traditional pill type detection method in the prior art cannot accurately detect the type of pills due to the influence of the external environment.

[0007] The technical solution of the present invention is: A white pill recognition system based on a short-wave infrared focal plane detector, comprising: Infrared light source: emits short-wave infrared light; Short-wave infrared detector: used to detect and sense infrared light radiation; Focusing module: arranged between the tablet and the short-wave infrared detector, when the infrared light penetrates the surface material of the tablet and is reflected, the reflected infrared light is focused onto the short-wave infrared detector; Main control module: used to drive the short-wave infrared detector to generate pixel current, and perform algorithm processing on the digital signal to form a frame; Data processing module: convert analog signals into digital signals; Display module: displays the framed image processed by the main control module.

[0008] Preferably, the infrared light emitted by the infrared light source is short-wave infrared light, and the wavelength of the infrared light is 1450nm-1650nm.

[0009] Preferably, the main control module includes a timing generation circuit, an external clock, and a bias current, the bias current adjusts or sets a specific current value, and the external clock drives the timing generation circuit to generate a timing pulse.

[0010] Preferably, the external clock includes a main clock and an integral control clock, the main clock is a synchronization reference, the short-wave infrared detector generates a timing pulse, the shortest high-level pulse width of the integral control clock is 3075 cycles, and when the integral control clock is at a high level, the MOS tube in the short-wave infrared detector is turned on, and when the integral control clock moves to the falling edge, the MOS tube is turned off and the signal integration ends.

[0011] Preferably, when the data processing module transmits and collects pixels on the short-wave infrared detector, between two adjacent frame integrations, there is at least one master clock interval between the falling edge of the output synchronization signal and the rising edge of the integration control clock.

[0012] Preferably, the short-wave infrared detector is exposed in a snapshot manner, and when the data processing module integrates the pixels, the pixel array starts integration and ends integration at the same time.

[0013] Preferably, the pixel size of the planar photovoltaic quantum dot short-wave infrared imaging chip in the short-wave infrared detector is 640×512.

[0014] A white pill recognition method based on a short-wave infrared focal plane detector, characterized in that it comprises the following steps: S1: Emit infrared light: The infrared light source emits infrared light to the tablet, and the infrared light penetrates the surface material of the tablet and is reflected; S2: receiving infrared light: the infrared light reflected in step S1 is focused on the short-wave infrared detector planar photovoltaic quantum dot short-wave infrared imaging chip through a focusing module; S3: Analog signal generation: The main control module drives the short-wave infrared detector to generate pixel current, and converts it into an analog signal through the circuit inside the short-wave infrared detector; S4: Data acquisition and analog-to-digital conversion: The buffer circuit in the data processing module receives the analog signal and transmits the analog signal to the four-channel analog-to-digital conversion circuit to convert the analog signal into a digital signal; S5: Data processing and image display: The main control module receives the digital signal converted in step S4, processes the digital signal by algorithm and maps it into a pixel matrix, and displays the image in the display module.

[0015] Preferably, step S5 includes the following sub-steps: S5.1: Convert electrical signals into grayscale images: Map the electrical signal of each pixel to a grayscale level to represent the brightness of the pixel, thereby converting the original electrical signal data into a displayable image form; S5.2: Grayscale image converted to grayscale histogram: The grayscale value of the grayscale image in S5.1 is obtained through the function in OpenCV, and compared with the preset threshold to determine whether the target exists, the number of times each pixel appears is counted, and drawn into a grayscale histogram.

[0016] Preferably, in step S5, the main control module extracts the shape and grayscale value of the pills through image processing technology, and uses a deep neural network to train a large amount of image data obtained from the short-wave infrared detector so that the network learns to recognize different types of pills.

[0017] Compared with the prior art, the advantages of the present invention are: (1) Short-wave infrared light has strong penetrability and can penetrate the surface material of the tablet to perform detection quickly and accurately; (2) Short-wave infrared light can be adjusted to different bands. Short-wave infrared light of different bands has different reflection and absorption characteristics for different types of tablets, which can improve the accuracy of detection of different types of tablets; (3) The signal strength received by the short-wave infrared detector is enhanced by the focusing module, thereby improving the sensitivity and accuracy of detection, so that even tablets with weak reflection signals can be effectively detected; (4) By comparing the similarities and differences between the grayscale histograms of two tablets, different types of tablets can be accurately distinguished. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The present invention will be further described below in conjunction with the accompanying drawings and embodiments: Figure 1 This is a flow chart of a white pill recognition method based on a short-wave infrared focal plane detector according to the present invention; Figure 2 This is a flow chart of step S5 of the present invention; Figure 3 This is a partial structural schematic diagram of a white pill recognition system based on a short-wave infrared focal plane detector according to the present invention; Figure 4 A comparison chart of the results of drug testing by the present invention; Figure 5Photographs of results of identifying the tablets for the present invention; Figure 6 A grayscale histogram generated by the present invention for the result photo; Figure 7 The white pill recognition system based on the short-wave infrared focal plane detector of the present invention; Figure 8 The white pill recognition device based on the short-wave infrared focal plane detector of the present invention; Fig. 9 The following are the results of the recognition system of the present invention on different tablets. DETAILED DESCRIPTION

[0019] The present invention is further described in detail below in conjunction with specific embodiments: like Figure 1-Figure 3 As shown, a white pill recognition system based on a short-wave infrared focal plane detector includes an infrared light source, a short-wave infrared detector, a focusing module, a main control module, a data processing module, and a display module; the infrared light source emits infrared light to illuminate the pill, and the infrared light is reflected after passing through the surface of the pill, and the reflected infrared light is focused on the short-wave infrared detector through the focusing module, the main control module drives the short-wave infrared detector to generate pixel current and converts it into an analog signal, the data processing module collects the analog signal and converts the analog signal into a digital signal, and the main control module converts the analog signal into an image through image algorithm processing after receiving the digital signal, and displays it on the display module; A white pill recognition method based on a short-wave infrared focal plane detector comprises the following steps: S1: Emit infrared light: The infrared light source emits short-wave infrared light with an adjustable wavelength within 1450nm-1650nm. The short-wave infrared light irradiates the tablet, penetrates the substance on the surface of the tablet and reflects; Among them, short-wave infrared light has strong penetrability and can penetrate the surface material of the tablets for rapid and accurate detection. In addition, short-wave infrared light has high spatial resolution and detection accuracy, and can accurately identify subtle features of tablets and slight differences between different batches. It can operate within different wavelength ranges. Secondly, short-wave infrared light of different bands has different reflection and absorption characteristics for different types of tablets, which can improve the accuracy of detection of different types of tablets.

[0020] Furthermore, the short-wave infrared light in the 1450nm-1650nm band has a low absorption rate for air and water, which enables the short-wave infrared light in this band to better penetrate the surface of the tablets. In most environmental conditions, the short-wave infrared light in this band can better reduce background interference, and can more clearly capture the features and details of the tablet surface during detection, thereby improving the accuracy and stability of detection. At the same time, short-wave infrared light in the 1450nm-1650nm band has high sensitivity to common chemical components in tablets, such as but not limited to different drug ingredients or additives, and can help determine the quality and consistency of drugs. Infrared light in this band is generally easier to achieve efficient optical performance. The optical elements and detectors of these systems in this band are relatively easy to obtain, which is beneficial to the stability and reliability of the system.

[0021] S2: receiving infrared light: the infrared light reflected in step S1 is focused on the short-wave infrared detector planar photovoltaic quantum dot short-wave infrared imaging chip through a focusing module; Specifically, the focusing module uses an optical lens. The focusing effect of the optical lens can focus the infrared light reflected from the surface of the tablet onto the short-wave infrared detector, thereby enhancing the signal strength received by the short-wave infrared detector and improving the sensitivity and accuracy of the detection. Even tablets with weak reflection signals can be effectively detected. At the same time, the optical lens can distinguish the infrared light reflected from different areas of the tablet, effectively distinguish the tiny structures, color differences or other characteristic changes on the surface of the tablet, improve the accuracy of detection of different types of tablets, and more effectively separate the differences between the surface features of the tablet and the background by precisely focusing on the infrared light reflected on the surface of the tablet, thereby improving the reliability and stability of the detection.

[0022] Specifically, the imaging chip has a pixel size of 640×512. Too low a pixel size means that the imaging resolution is reduced, and subtle changes and target details cannot be captured, which will make it difficult to identify the type of tablets. It will also lead to insufficient data sampling, resulting in blurred or lack of details in the imaging results, thereby reducing the accuracy of the detection. If the pixel size is too large, higher processing power and storage space are required to effectively process and store data, which will increase the complexity and cost of system design.

[0023] S3: Analog signal generation: The main control module drives the short-wave infrared detector to generate pixel current, and converts it into an analog signal through the circuit inside the short-wave infrared detector; The main control drives the detector to convert the collected optical signal into an analog signal through the detector circuit and then transmit it to the buffer circuit; S4: Data acquisition and analog-to-digital conversion: The buffer circuit in the data processing module receives the analog signal and transmits the analog signal to the four-channel analog-to-digital conversion circuit to convert the analog signal into a digital signal; S5: Data processing and image display: When the buffer circuit completes the data collection conditions, the analog signal in the buffer circuit will be transmitted to the four-way analog-to-digital conversion circuit to convert the analog signal into a digital signal, and the digital signal to be processed will be input into the main control module. The main control module processes the input data according to the data processing algorithm of digital filtering, signal enhancement and feature extraction. The processed data will be mapped to the pixel matrix, and each pixel corresponds to a set of parameters of the digital signal, and finally displayed on the display module through the serial port.

[0024] Furthermore, the main control chip in the main control module processes the obtained digital signal through an algorithm, frames it into a picture and displays it on the display module. The shape and grayscale value of the white tablet are extracted through image processing technologies such as edge detection and binarization processing. A deep neural network is used to train a large amount of image data obtained from the short-wave infrared focal plane detector, so that the network can learn to recognize different types of tablets, thereby realizing the recognition of tablet types.

[0025] S5.1: Convert electrical signals into grayscale images: Map the electrical signal of each pixel to a grayscale level to represent the brightness of the pixel, thereby converting the original electrical signal data into a displayable image form; S5.2: Grayscale image converted to grayscale histogram: The grayscale value of the grayscale image in S5.1 is obtained through the function in OpenCV, and compared with the preset threshold to determine whether the target exists, the number of times each pixel appears is counted, and drawn into a grayscale histogram; Furthermore, the grayscale histogram can be used to identify features such as brightness and texture of the comparison image. When recognizing an image, the grayscale histogram provides a global description of the image content. By comparing the grayscale histograms of different images, their similarities or differences can be evaluated, and the types of pills can be accurately and efficiently identified.

[0026] In this embodiment, Figure 4 As shown, the raw data output by the shortwave infrared detector is a set of continuous electrical signals, which represent the intensity of infrared radiation received by each pixel on the shortwave infrared detector. In order to convert these electrical signals into visible images, they need to be converted into grayscale images. A grayscale image is an image with only a single brightness channel. It represents the brightness of each pixel by mapping the electrical signal of each pixel to a grayscale level, converting the raw electrical signal data into an image form that is easier to understand and display. In addition, grayscale images have higher contrast, which facilitates better observation and analysis of details and features in infrared images.

[0027] Specifically, Figure 5 As shown, statistical analysis and feature extraction are performed on the grayscale image to generate a grayscale histogram, which can describe the distribution of pixel intensity and present it in the form of frequency or probability of pixel intensity changes.

[0028] Furthermore, the main control module includes a timing generation circuit, an external clock, and a bias current. The timing generation circuit uses the external clock and the bias current to generate the timing pulse voltage required for the detector to work; the external clock includes a master clock and an integral control clock. The short-wave infrared detector uses the master clock as a synchronization reference. The timing generation circuit generates timing pulses based on the master clock. The integral control clock controls the exposure integration and signal readout of the detector; during the high-level time of the integral control clock, the MOS tube in the detector is turned on, and the photogenerated current is stored on the integral capacitor. Within 3075 cycles after the rising edge of the integral control clock, the short-wave infrared detector resets the integral capacitor, and only then does it start the effective photogenerated signal integration. Therefore, the shortest high-level pulse width of the integral control clock is 3075 cycles. On this basis, the high-level pulse width of the integral control clock can be arbitrarily changed according to application needs. After the falling edge of the integral control clock arrives, the injected MOS tube is turned off to end the signal integration.

[0029] Furthermore, after the integration is completed, the data processing module starts to transmit and sample the first row of pixels in the start window of the short-wave infrared detector. After that, the continuous output of the array pixel signal is realized through the ping-pong structure. While the array pixel signal is read out, the output synchronization signal is converted to a high level for external circuit sampling synchronization. After the array pixel signal is read out, the output synchronization signal is converted to a low level. Before starting the next frame of integration, there must be at least one main clock between the falling edge of the output synchronization signal and the rising edge of the integration control clock to ensure timing consistency and stability. By leaving enough time interval between the signal change and the clock edge, competition conditions can be avoided to ensure that the input signal is stable when the clock edge arrives. The short-wave infrared detector is exposed in a snapshot manner, and the pixel array starts integration at the same time and ends integration at the same time.

[0030] The working principle of this embodiment: When detecting the type of tablets, the infrared light source emits short-wave infrared light and irradiates it on the tablets. The short-wave infrared light penetrates the material on the surface of the tablets and reflects. The reflected infrared light passes through the focusing module, and the focusing module focuses the reflected light on the short-wave infrared detector. At this time, the main control module drives the short-wave infrared detector to generate pixel current, and the data conversion module collects the pixel current and converts it into a digital signal. The main control module processes the obtained digital signal through an image algorithm to obtain a grayscale image and a grayscale histogram and display them on the display module. The type of tablets can be identified by comparing the grayscale histograms of the two tablets.

[0031] The above embodiments are only for illustrating the technical concept and features of the present invention, and their purpose is to enable people familiar with this technology to understand the content of the present invention and implement it accordingly, and they cannot be used to limit the scope of protection of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is limited by the attached claims rather than the above description, and it is intended to include all changes within the meaning and scope of the equivalent elements of the claims in the present invention.

Claims

1. A white pill recognition system based on short-wave infrared focal plane detector, characterized in that: include: Infrared light source: emits short-wave infrared light; Short-wave infrared detector: used to detect and sense infrared light radiation; Focusing module: arranged between the tablet and the short-wave infrared detector, when the infrared light penetrates the surface material of the tablet and is reflected, the reflected infrared light is focused onto the short-wave infrared detector; Main control module: used to drive the short-wave infrared detector to generate pixel current, and perform algorithm processing on the digital signal to form a frame; Data processing module: convert analog signals into digital signals; Display module: displays the framed image processed by the main control module.

2. The white pill recognition system based on short-wave infrared focal plane detector according to claim 1 is characterized in that: The infrared light emitted by the infrared light source is short-wave infrared light, and the wavelength of the infrared light is 1450nm-1650nm.

3. The white pill recognition system based on short-wave infrared focal plane detector according to claim 1 is characterized in that: The main control module includes a timing generation circuit, an external clock, and a bias current. The bias current adjusts or sets a specific current value. The external clock drives the timing generation circuit to generate a timing pulse.

4. The white pill recognition system based on short-wave infrared focal plane detector according to claim 3 is characterized in that: The external clock includes a main clock and an integral control clock, the main clock is a synchronization reference, the short-wave infrared detector generates a timing pulse, the shortest high-level pulse width of the integral control clock is 3075 cycles, and when the integral control clock is at a high level, the MOS tube in the short-wave infrared detector is turned on, and when the integral control clock moves to the falling edge, the MOS tube is turned off and the signal integration ends.

5. The white pill recognition system based on short-wave infrared focal plane detector according to claim 4 is characterized in that: When the data processing module transmits and collects pixels on the short-wave infrared detector, between two adjacent frame integrations, there is at least one master clock interval between the falling edge of the output synchronization signal and the rising edge of the integration control clock.

6. The white pill recognition system based on short-wave infrared focal plane detector according to claim 1 is characterized in that: The short-wave infrared detector is exposed in a snapshot manner. When the data processing module integrates pixels, the pixel array starts integration and ends integration at the same time.

7. The white pill recognition system based on short-wave infrared focal plane detector according to claim 1 is characterized in that: The pixel size of the planar photovoltaic quantum dot short-wave infrared imaging chip in the short-wave infrared detector is 640×512.

8. A method for identifying white pills based on a short-wave infrared focal plane detector, implemented by a white pill identification system based on a short-wave infrared focal plane detector according to any one of claims 1 to 7, characterized in that: The following steps are involved: S1: Emit infrared light: The infrared light source emits infrared light to the tablet, and the infrared light penetrates the surface material of the tablet and is reflected; S2: receiving infrared light: the infrared light reflected in step S1 is focused on the short-wave infrared detector planar photovoltaic quantum dot short-wave infrared imaging chip through a focusing module; S3: Analog signal generation: The main control module drives the short-wave infrared detector to generate pixel current, and converts it into an analog signal through the circuit inside the short-wave infrared detector; S4: Data acquisition and analog-to-digital conversion: The buffer circuit in the data processing module receives the analog signal and transmits the analog signal to the four-channel analog-to-digital conversion circuit to convert the analog signal into a digital signal; S5: Data processing and image display: The main control module receives the digital signal converted in step S4, processes the digital signal by algorithm and maps it into a pixel matrix, and displays the image in the display module.

9. The method for identifying white pills based on a short-wave infrared focal plane detector according to claim 8, characterized in that: Step S5 includes the following sub-steps: S5.1: Convert electrical signals into grayscale images: Map the electrical signal of each pixel to a grayscale level to represent the brightness of the pixel, thereby converting the original electrical signal data into a displayable image form; S5.2: Grayscale image converted to grayscale histogram: The grayscale value of the grayscale image in S5.1 is obtained through the function in OpenCV, and compared with the preset threshold to determine whether the target exists, the number of times each pixel appears is counted, and drawn into a grayscale histogram.

10. The method for identifying white pills based on a short-wave infrared focal plane detector according to claim 8, characterized in that: In step S5, the main control module extracts the shape and grayscale value of the pills through image processing technology, and uses a deep neural network to train a large amount of image data obtained from the short-wave infrared detector so that the network learns to recognize different types of pills.

Citation Information

Cited By

  • Sensing focal plane image recognition system based on FPGA acceleration

    CN120852724A

  • An FPGA-accelerated focal plane image recognition system

    CN120852724B