Monocular vision measurement system and method based on dual-core heterogeneous architecture

Abstract

The invention belongs to the crossing field of computer vision and embedded systems, and particularly relates to a monocular vision measurement system and method based on a dual-core heterogeneous architecture. The invention discloses a monocular vision measurement system based on a dual-core heterogeneous architecture. An image acquisition module acquires a monocular image; the visual processing unit executes target detection, contour extraction, perspective transformation and a size measurement algorithm, and the visual processing unit receives the monocular image and outputs a target measurement result; the current detection module is electrically connected with the control unit and is used for monitoring system power consumption in real time and outputting a proportional voltage signal; the control unit is used for system scheduling, current signal acquisition and communication control, receives a target measurement result and a proportional voltage signal input by the current detection module, and outputs the target measurement result and the proportional voltage signal to the man-machine interaction module; and the man-machine interaction module is used for displaying a measurement result and a system state. Through heterogeneous design of the control unit and the visual processing unit, effective cooperation of high-precision visual processing and low-power-consumption real-time control is realized.

Description

TECHNICAL FIELD

[0001] The present application belongs to the cross field of computer vision and embedded systems, and particularly relates to a monocular vision measurement system and method based on a dual-core heterogeneous architecture, which is especially suitable for real-time target size and distance measurement in a resource-limited environment. BACKGROUND

[0002] With the rapid development of industrial automation and intelligent warehousing, the demand for low-cost and high-efficiency non-contact measurement technology is increasingly urgent. Traditional contact measurement methods have problems such as low efficiency and easy damage to the measured object. In the non-contact measurement scheme, it is difficult to balance low cost and high precision. In recent years, especially the non-contact measurement technology based on machine vision, because of its high precision, high speed, non-destructive testing and other characteristics, it has gradually become a research hotspot in the field of industrial measurement.

[0003] Monocular vision measurement, as an important branch of computer vision, simulates human visual perception with a single camera, combines digital image processing and algorithm analysis, and realizes the identification, positioning and measurement of target objects. Compared with stereo vision systems, monocular vision systems have obvious advantages such as simpler hardware structure, lower cost and reduced computational complexity, and are more suitable for large-scale deployment and application in industrial environments.

[0004] Monocular vision measurement mainly includes the following three ways in terms of algorithm and calculation: (1) The measurement system based on traditional geometric vision adopts classic image processing algorithms (such as edge detection, contour extraction, perspective transformation, etc.), combines camera calibration parameters, and calculates target size and distance through similar triangle principle or PnP (Perspective-n-point) algorithm. Its main dependence on manually designed feature extraction algorithms, in complex scenes such as light changes, target texture missing, occlusion, etc., the stability is insufficient, the environmental robustness is poor; and there are defects such as limited real-time performance and insufficient flexibility. (2) The end-to-end measurement system based on deep learning generally uses convolutional neural networks or Transformer models to directly regress the geometric parameters of the target from the image. Its main defects are: the contradiction between computing power and power consumption is prominent, the GPU / NPU platform has high power consumption, which is difficult to deploy for a long time in battery-powered or strictly limited heat dissipation scenarios; and a large amount of accurately labeled training data is required, which has high labeling cost and limited generalization ability. (3) The lightweight vision system based on single-core microcontrollers runs simplified version of vision algorithms (such as binarization, template matching) on STM32 microcontrollers, etc. Due to the low frequency of microcontrollers and the lack of dedicated vision acceleration units, it is difficult to process complex algorithms, and it is difficult to meet the needs of dynamic scenes in terms of measurement accuracy and frame rate. In order to adapt to the computing power, the algorithm integrity is often sacrificed, resulting in large measurement errors, which is only suitable for extremely simple fixed scenes. SUMMARY

[0005] The application aims to overcome the deficiencies in the prior art, and provide a monocular vision measurement system and method based on a dual-core heterogeneous architecture, which adopts a heterogeneous design of a control unit + vision processing unit to realize the cooperation of high-performance vision processing and low-power real-time control.

[0006] To solve the above technical problems, the application discloses a monocular vision measurement system based on a dual-core heterogeneous architecture, which comprises an image acquisition module, a vision processing unit, a current detection module, a control unit and a human-computer interaction module, and realizes the effective cooperation of high-precision vision processing and low-power real-time control. It is particularly suitable for resource-limited industrial scenes. The system adopts a self-developed algorithm to optimize the monocular ranging process through minimum edge length extraction and perspective correction, significantly enhancing the robustness and precision of complex target measurement.

[0007] The image acquisition module is used to acquire monocular images containing standard reference objects and target objects; the vision processing unit is used to execute target detection, contour extraction, perspective transformation and size measurement algorithms, receives the monocular images acquired by the image acquisition module and outputs target measurement results; the current detection module is electrically connected with the control unit and is used to monitor the power consumption of the system in real time and output a proportional voltage signal; the control unit is used for system scheduling, current signal acquisition and communication control, receives the target measurement results input by the vision processing unit and the proportional voltage signal input by the current detection module, and outputs them to the human-computer interaction module; and the human-computer interaction module is used to display measurement results and system states.

[0008] Preferably, the standard reference object is A4 paper, the target object is a rectangular image on the A4 paper, and the algorithm executed by the vision processing unit comprises the following steps: detecting the rectangular contour of the input image, and forming a candidate rectangular set from multiple candidate rectangles meeting the rectangular detection condition; verifying the candidate rectangles in the candidate rectangular set according to the order from large to small in area, defining the candidate rectangle with the largest area meeting the standard proportion as a candidate reference rectangle, and correcting the candidate reference rectangle through perspective transformation to generate a standard reference rectangle without distortion; extracting a target rectangle from the candidate reference rectangle in the standard reference rectangle, wherein the target rectangle is derived from the candidate rectangles other than the standard reference rectangle in the candidate rectangular set, and the target rectangle is completely located inside the standard reference rectangle; selecting an area inside the standard reference rectangle outside the target rectangle as an ROI area, performing binaryzation processing on the image in the ROI area, and performing Blob detection on the binaryzation image to generate multiple connected pixel areas, defining the connected pixel areas as feature spots, and performing shape recognition and geometric shape verification on each feature spot, and defining the feature spot meeting the requirements as the target object.

[0009] Based on the known physical size and pixel size of the standard reference rectangle, the actual geometric parameters of the target are calculated according to the pixel size of the target object in the image.

[0010] More preferably, all the rectangular contours in the image are detected based on edge detection and polygon approximation method, and a candidate rectangle set is formed, each rectangular contour is reversely mapped into an orthographic view based on a perspective geometric model, the aspect ratio R of the rectangular orthographic view is calculated, and the rectangular orthographic view with the aspect ratio R asp that is closest to the aspect ratio of A4 paper is selected as the standard reference rectangle. asp

[0011] Preferably, the aspect ratio R asp of the selected rectangle satisfies formula (1), and the candidate rectangle is determined as the standard reference rectangle. (1), wherein, is the aspect ratio of A4 paper, is a threshold value set by human, representing the allowed skew range.

[0012] Preferably, the screening of the candidate reference rectangle includes: candidate rectangle set sorting: the candidate rectangles in the candidate rectangle set are arranged in descending order of area, reference verification: starting from the candidate rectangle with the largest area, the aspect ratio of the candidate rectangle is verified, reference confirmation: the first candidate rectangle that passes the aspect ratio verification is determined as the candidate reference rectangle, target rectangle searching: in the remaining candidate rectangles, it is searched whether there is a target rectangle completely located inside the candidate reference rectangle, if yes, the candidate reference rectangle and the target rectangle are matched, and the process ends, if no, the process enters a rollback process; rollback process: the current candidate reference rectangle is abandoned, the next candidate rectangle that passes the aspect ratio verification is taken in order of area, and the target rectangle searching step is repeated until the target rectangle is successfully found or all the candidate rectangles cannot find the target rectangle inside, and the process ends.

[0013] Preferably, the quadrilateral recognition is performed on each feature spot by using the minimum bounding rectangle corner method, and the feature spot that meets the definition of the quadrilateral is defined as the candidate feature spot.

[0014] The internal filling rate R Area of the candidate feature spot is determined to determine the geometric shape of the candidate feature spot, and the internal filling rate R Area is calculated in the manner shown in formula (2), (2), wherein A​​blob A represents the area of ​​all pixels in the candidate feature blob. minrect Let R be the area of ​​the minimum bounding rectangle corresponding to the candidate feature blob, when the internal fill rate R Area When the fill rate is greater than the threshold, the candidate feature blob is identified as the target object.

[0015] Based on the actual physical size of A4 paper and its pixel size in the image, a scaling factor is obtained. Then, the actual physical size of each target object is the product of its pixel size and the scaling factor, where the minimum side pixel length is used as the feature width.

[0016] The distance D from the target object to the camera is calculated based on the principle of similar triangles under the pinhole camera model. The calculation method is shown in formula (3): (3), among which, This refers to the actual physical width of an A4 sheet of paper. Let be the equivalent focal length of the camera, which is known through camera calibration. The pixel width of the reference object in the image.

[0017] Preferably, the internal fill rate threshold is 0.9.

[0018] Preferably, the current detection module includes a signal conditioning circuit based on an operational amplifier, which amplifies and filters the voltage difference across the sampling resistor before inputting it to the ADC interface of the control unit; the human-machine interaction module is a serial touch screen, which communicates with the control unit via the UART protocol to realize real-time display of measurement data and reception of user commands; the vision processing unit is a K210 or K230 chip, and the control unit is an STM32 series microcontroller, which interact with the control unit via a UART or SPI interface for data and command exchange.

[0019] On the other hand, this invention discloses a monocular vision measurement method based on a dual-core heterogeneous architecture, comprising the following steps: The image acquisition module acquires a monocular image containing both a standard reference object and the target object. The vision processing unit identifies standard reference objects and performs perspective correction on them, establishing a mapping relationship between image pixels and actual physical dimensions; Identify the target object within a standard reference and extract its pixel dimensions; Based on the mapping relationship and the pixel size of the target, the actual physical size of the target and its distance from the camera are calculated; The control unit synchronously collects the system power consumption and sends the measurement results to the human-machine interaction module for display.

[0020] Preferably, it also includes system power management steps, such as reducing the operating frequency of the vision processing unit when it is idle, and enabling DMA data transfer in the control unit.

[0021] The standard reference object is a rectangle that conforms to the aspect ratio of A4 paper, and the target object is a rectangle that is completely located within the standard reference object and has a fill rate greater than the fill rate threshold. During the identification of the standard reference object, if there are multiple rectangular outlines in the image, they are sorted from largest to smallest area, and proportional verification and inclusion relationship verification are performed in sequence to determine the unique outer frame of the standard reference object and the inner frame of the target object.

[0022] The beneficial effects of this invention are as follows: 1. A dual-core heterogeneous architecture combining an STM32 series microprocessor and a K230 vision processing unit is adopted, achieving efficient task decoupling and resource optimization. The control unit handles high-real-time, computationally simple tasks such as high-precision current acquisition, system scheduling, and communication control, while the vision processing unit is dedicated to image target detection, perspective transformation, and size calculation. Its NPU accelerates neural networks and image algorithms, avoiding single-core overload and significantly improving system real-time performance. Furthermore, the STM32 uses DMA transfer to reduce CPU load and dynamic power consumption, while the K230 supports dynamic frequency adjustment to reduce static power consumption, achieving a balance between low power consumption and high energy efficiency.

[0023] 2. Measurement accuracy and robustness are improved. Adaptive geometric correction based on A4 reference is used, and a certain range of skew and perspective distortion errors are allowed. The minimum side pixel length of the quadrilateral is calculated as the feature width to reduce the impact of perspective distortion on size transformation. When determining the target object, the A4 outer frame is located first, and then the target is detected in the ROI inside it to eliminate background interference.

[0024] 3. The system adopts a modular design, making it easy to deploy. The system is divided into an image acquisition module, a vision processing unit, a current detection module, a control unit, and a human-machine interface module. The interfaces are clear and easy to customize and expand. The current detection circuit is based on operational amplifier amplification and filtering, and STM32 ADC acquisition, monitoring system power consumption in real time and providing data support for energy optimization and fault diagnosis.

[0025] 4. Lower cost and greater adaptability. Compared to binocular or multi-camera systems, this invention has lower hardware costs, making it suitable for large-scale industrial deployment; and its algorithm is lightweight, allowing it to run in real time on embedded edge devices. Attached Figure Description

[0026] Figure 1 This is an overall structural diagram of the system of the present invention; Figure 2 This is a flowchart illustrating the process of determining the target rectangle in this invention; Figure 3 This is a schematic diagram of the process for determining the candidate reference rectangle in this invention; Figure 4 This is a circuit diagram of the current detection module in this invention; Figure 5 This is a structural diagram of the detected image in this invention; Figure 6 This is the result image of the detection. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0028] like Figure 1 As shown, a monocular vision measurement system based on a dual-core heterogeneous architecture includes an image acquisition module, a vision processing unit, a current detection module, a control unit, and a human-computer interaction module, realizing the process from data acquisition to vision processing to control decision-making to human-computer interaction.

[0029] The image acquisition module uses a USB network camera with a resolution of 1920*1080 and a frame rate of 30 FPS. The camera is connected to the vision processing unit K230 via a USB interface and is responsible for acquiring monocular images containing reference objects and the target object to be measured.

[0030] The vision processing unit uses the Canaan K230 edge AI chip, which has a built-in dual-core NPU with a computing power of about 6 TOPs, meeting the real-time requirements of neural network inference and image processing. The K230 communicates with the control unit STM32 through the UART interface to output measurement results such as target distance and size.

[0031] The current detection module is based on the LM358 dual operational amplifier design for current sampling circuit design, such as... Figure 4 As shown in the diagram, the sampling current flows through sampling resistor R8, generating a voltage drop. This voltage drop signal is sent to the inverting input of operational amplifier LM358 U1.1 via resistor R5. U1.1 amplifies the signal and outputs it to R1, then feeds it back to the inverting input of U1.1 via R3, forming a closed-loop control and enhancing circuit stability. U1.2 amplifies the output of U1.1 twice and outputs it to R2. Finally, it is filtered by capacitor C2 to obtain a stable ADC1 signal. Capacitors C1, C3, C5, C6, and C4 are used in the circuit to stabilize the power supply and signal, reducing interference. This circuit achieves precise amplification and filtering of the current signal, outputting a stable analog signal for ADC sampling, and is suitable for applications requiring precise current measurement.

[0032] The control unit uses an STM32F103C8T6 microcontroller as its control unit. It is responsible for acquiring the analog voltage signal output by the current detection module through the ADC channel, receiving the measurement data sent by the K230 through the UART, sending the measurement data to the human-machine interaction module through another UART interface, and system task scheduling and communication protocol processing.

[0033] The human-computer interaction module uses a 7-inch UART serial port touch screen. The display screen is connected to the STM32 via UART to display the target distance, size, system power consumption and status information in real time, and supports touch command input.

[0034] In addition, the entire system is powered by an external 5V / 3A DC power supply, which is regulated to 3.3V by an LDO to supply the STM32 and K230 core circuits.

[0035] like Figure 2 As shown, the process by which this system determines the target object is as follows: Phase 1: Image Acquisition and Preprocessing: The image acquisition module acquires images and converts them into the input format for the visual processing unit.

[0036] Phase 2: Rectangular Contour Detection and Candidate Set Generation: The `img.find_rects` function is called to extract all quadrilateral contours in the image based on Canny edge detection and Douglas-Peucker polygon approximation. Each quadrilateral undergoes an inverse perspective transformation and is added to the candidate rectangle set.

[0037] Phase 3: Reference Object and Target Rectangle Selection: such as Figure 3 As shown, the candidate rectangle set is sorted in descending order of area. Starting with the rectangle with the largest area, the rectangle is verified one by one to see if it passes the A4 scale verification. The first rectangle that passes the verification is defined as the candidate reference rectangle. The remaining rectangles are searched to see if there is a rectangle that is completely inside the candidate reference rectangle. If found, it is defined as the target rectangle. Otherwise, the backtracking process is activated: the current candidate reference rectangle is abandoned, and the next rectangle that passes the verification is selected and the search is repeated.

[0038] In this stage, A4 scale verification allows for errors as long as the aspect ratio R of the candidate rectangle's front view is within acceptable limits. asp satisfy (in If the ratio is 0.05, it can be determined that it meets the A4 paper ratio verification.

[0039] Phase 4: Perspective Transformation and ROI Extraction: In the matched candidate reference rectangle and target rectangle view, the ROI region is defined as the area 10 pixels outside the target rectangle. The ROI region is located inside the candidate reference rectangle.

[0040] Phase 5: Blob Detection and Feature Extraction: Convert the ROI region to grayscale, perform binarization (threshold 128), conduct connectivity component analysis, and extract blobs with an area between 100 and 100,000 pixels. For each blob, calculate its minimum bounding rectangle corner points and its fill rate. ,like If the value is greater than 0.9, it is considered a valid target feature spot.

[0041] Phase 6: Size and Distance Calculation: First, the actual width of the A4 paper is known. =210mm, the pixel width in the image is ,and The candidate reference rectangle is selected based on its minimum side length after perspective transformation. Choosing the minimum side length enhances the robustness of the algorithm. The scaling factor is then used. The actual size of each target feature spot is obtained by multiplying the pixel size of its smallest bounding rectangle by the scaling factor k.

[0042] The distance D is calculated based on the pinhole model formula. ,in This can be obtained through camera calibration.

[0043] Phase 7: Data Output and Display: The K230 sends the target distance, size, and coordinate information to the STM32 via UART. The STM32 simultaneously collects current and voltage signals, calculates real-time power consumption, and packages all data to be displayed on the serial port screen.

[0044] Phase 8: Power Management: During the idle period of K230, the frequency reduction API is automatically called to reduce the main frequency to 500MHz, while STM32 enables DMA to transfer ADC data, reducing CPU utilization.

[0045] The pseudocode for this algorithm is shown in Table 1: Table 1 Pseudocode of Target Recognition Algorithm This invention also discloses a monocular vision measurement method based on a dual-core heterogeneous architecture, comprising the following steps: acquiring a monocular image containing a standard reference object and a target object through an image acquisition module; identifying the standard reference object and performing perspective correction on it to establish a mapping relationship between image pixels and actual physical dimensions; identifying the target object in the same image and extracting its pixel dimensions; calculating the actual physical dimensions of the target and its distance from the camera based on the mapping relationship and the pixel dimensions of the target; and synchronously acquiring system power consumption by a control unit and sending the measurement results to a human-computer interaction module for display.

[0046] The standard reference is an A4 sheet of paper, and the target object is a rectangular image that is completely located within the standard reference and has a fill rate greater than the fill rate threshold. During the identification of the standard reference, if there are multiple rectangular outlines in the image, they are sorted by area from largest to smallest, and proportional verification and inclusion relationship verification are performed in sequence to determine the unique outer frame of the standard reference and the inner frame of the target object.

[0047] like Figure 5 As shown, to verify system performance, four target types are compared: Figure 5 (a) represents the unilateral SS, Figure 5 (b) represents the square composite pattern SNL with a number label. Figure 5 The composite pattern MSS, consisting of multiple separate squares, and the composite pattern POS, consisting of partially overlapping squares, represented in (c), were tested five times using this system and method within a distance range of 115-155 cm.

[0048] The experimental results are as follows Figure 6 As shown in Table 2, the maximum error in distance measurement is ≤4.2%, and the maximum error in side length measurement is ≤4.8%. The system power consumption is shown in Table 2. The above experimental results show that the developed measuring device can control the real-time measurement error of the distance, side length, and area of ​​the moving target within ±5% while maintaining stable system power consumption, thus meeting the design requirements of practical applications.

[0049] Table 2 System power consumption test results In this embodiment, the reference object is A4 paper, and the target object is a rectangle with a high fill rate. The standard reference object can also be replaced with other objects of known size, and the target object is not limited to a rectangle with a high fill rate.

[0050] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A monocular vision measurement system based on dual-core heterogeneous architecture, comprising an image acquisition module, characterized in that: It also comprises a visual processing unit, a current detection module, a control unit and a human-computer interaction module, realizing the process from data collection-visual processing-control decision-human-computer interaction, The image acquisition module is used for acquiring monocular images containing standard reference objects and target objects; The visual processing unit is used for executing target detection, contour extraction, perspective transformation and size measurement algorithms, and receives monocular images collected by the image acquisition module and outputs target measurement results; The current detection module is electrically connected with the control unit, and is used for monitoring system power consumption in real time and outputting proportional voltage signals; The control unit is used for system scheduling, current signal acquisition and communication control, receives target measurement results input by the visual processing unit and proportional voltage signals input by the current detection module, and outputs them to the human-computer interaction module; The human-computer interaction module is used for displaying measurement results and system states.

2. The monocular vision measurement system based on dual-core heterogeneous architecture according to claim 1, characterized in that: The standard reference object is A4 paper, and the target object is a rectangular image on the A4 paper, and the size measurement algorithm executed by the visual processing unit comprises the following steps: Rectangular contour detection is performed on the input image, and a plurality of candidate rectangles meeting the rectangular detection condition form a candidate rectangle set; The candidate rectangles in the candidate rectangle set are subjected to perspective transformation correction, and are subjected to standard proportion verification in descending order of area, the candidate rectangle meeting the standard proportion and having the largest area is determined as a candidate reference rectangle, and a standard reference rectangle without distortion is generated for the candidate reference rectangle; A target rectangle is extracted in the standard reference rectangle, the target rectangle is derived from the candidate rectangles other than the standard reference rectangle in the candidate rectangle set, and the target rectangle is completely located inside the standard reference rectangle; Based on the target rectangle, an ROI region is defined outside the target rectangle and inside the standard reference rectangle, binary processing is performed on the image in the ROI region, and Blob detection is performed on the binary processed image to generate a plurality of connected pixel regions, which are defined as feature spots, shape recognition and geometric shape verification are performed on each feature spot, and the feature spot meeting the requirements is identified as the target object; Based on the known physical size and pixel size of the standard reference rectangle, the actual geometric parameters of the target object are calculated according to the pixel size of the target object in the image.

3. The monocular vision measurement system based on a dual-core heterogeneous architecture according to claim 2, characterized in that: all rectangular contours in the image are detected based on an edge detection and a polygon approximation method, and a candidate rectangle set is formed, each rectangular contour is inversely mapped to an orthographic view based on a perspective geometric model, The aspect ratio R of the rectangular front view is calculated asp The aspect ratio R is filtered out asp The standard reference rectangle is the aspect ratio of A4 paper.

4. The monocular vision measurement system based on dual-core heterogeneous architecture according to claim 3, characterized in that: The aspect ratio R of the candidate rectangle asp When the formula (1) is satisfied, the candidate rectangle is determined as a standard reference rectangle. （1）， wherein, is the aspect ratio of A4 paper, is a threshold value set by a human being, indicating the skew allowed within the range.

5. The monocular vision measurement system based on dual-core heterogeneous architecture according to claim 4, characterized in that: a candidate reference rectangle is screened, candidate rectangles in the candidate rectangle set are arranged in descending order of area, a reference object is verified: starting from the candidate rectangle with the largest area, whether the aspect ratio of the candidate rectangle meets the standard A4 paper proportion is verified, a reference object is confirmed: the first candidate rectangle passing the aspect ratio verification is determined as the candidate reference rectangle, a target rectangle is searched for: in the remaining candidate rectangles, whether there is a target rectangle completely located inside the candidate reference rectangle is searched for, if there is, the candidate reference rectangle and the target rectangle are matched, and the process ends, If not, go to fallback procedure; Fallback procedure: abandon the current candidate reference rectangle, take the next candidate rectangle which passed the aspect ratio verification in area order, repeat the target rectangle searching procedure until successfully find the target rectangle or all candidate rectangles cannot find the target rectangle inside, end the procedure.

6. The monocular vision measurement system based on dual-core heterogeneous architecture according to claim 2, characterized in that: Each feature spot is identified as a quadrilateral by the method of minimum bounding rectangle angle, and the definition of the quadrilateral that meets the requirement is a candidate feature spot, An internal fill rate R of the candidate feature spot is utilized Area The internal fill rate R of the candidate feature spot is determined Area The calculation method is shown in formula (2), (2), wherein A blob is the total pixel area of the candidate feature spot, A minrect is the area of the minimum bounding rectangle corresponding to the candidate feature spot, and when the internal fill rate R Area is greater than a fill rate threshold, the candidate feature spot is determined as the target object. A proportion factor is obtained based on the actual physical size of A4 paper and its pixel size in the image, and the actual physical size of each target object is the product of its pixel size and the proportion factor, wherein the minimum edge pixel length is used as the feature width, The distance D of the target object to the camera is calculated based on the similar triangle principle under the pinhole camera model, and the calculation method is shown in formula (3): (3), wherein, is the actual physical width of A4 paper, is the pixel representation of the camera focal length, which is known through camera calibration, is the pixel width of the reference object in the image.

7. The monocular vision measurement system based on dual-core heterogeneous architecture according to claim 6, characterized in that: Internal filling rate R Area The threshold value is 0.

9.

8. The binocular heterogenous architecture based monocular vision measurement system according to claim 5 or 7, characterized in that: The current detection module includes an operational amplifier-based signal conditioning circuit that amplifies and filters the voltage difference across the sampling resistor and inputs it to the ADC interface of the control unit; The human-computer interaction module is a serial touch screen that communicates with the control unit through the UART protocol to realize real-time display of measurement data and reception of user instructions; The vision processing unit is a K210 or K230 chip, and the control unit is an STM32 series microcontroller, which interact with each other through UART or SPI interface.

9. A measuring method using the monocular vision measurement system based on a dual-core heterogeneous architecture according to claim 8, characterized in that: The method comprises the following steps: An image containing a standard reference and a target object is obtained by an image acquisition module; A vision processing unit identifies the standard reference and performs perspective correction thereon to establish a mapping relationship between image pixels and actual physical size; The target object is identified in the standard reference, and its pixel size is extracted; Based on the mapping relationship and the pixel size of the target, the actual physical size of the target and the distance from the camera are calculated; The control unit synchronously acquires system power consumption and sends the measurement results to the human-computer interaction module for display.

10. The monocular vision measurement method based on dual-core heterogeneous architecture according to claim 9, characterized in that: It further comprises a system power consumption management step, which reduces the working frequency of the vision processing unit when it is idle and enables DMA data transmission in the control unit; The standard reference is an A4 paper, and the target object is a rectangular image completely located in the standard reference and having a fill rate greater than a fill rate threshold; In the identification process of the standard reference, if there are multiple rectangular contours in the image, they are sorted in descending order of area, and proportion verification and inclusion verification are performed in turn to determine the unique standard reference outer frame and target object inner frame.

Images