Image Acquisition Module Research Articles

CNN-MLP-Based Configurable Robotic Arm for Smart Agriculture

Amidst escalating global populations and dwindling arable lands, enhancing agricultural productivity and sustainability is imperative. Addressing the inefficiencies of traditional agriculture, which struggles to meet the demands of large-scale production, this paper introduces a highly configurable smart agricultural robotic arm system (CARA), engineered using convolutional neural networks and multilayer perceptron. CARA integrates a highly configurable robotic arm, an image acquisition module, and a deep processing center, embodying the convergence of advanced robotics and artificial intelligence to facilitate precise and efficient agricultural tasks including harvesting, pesticide application, and crop inspection. Rigorous experimental validations confirm that the system significantly enhances operational efficiency, adapts seamlessly to diverse agricultural contexts, and bolsters the precision and sustainability of farming practices. This study not only underscores the vital role of intelligent automation in modern agriculture but also sets a precedent for future agricultural innovations.

PIAA: Pre-imaging all-round assistant for digital radiography

BACKGROUND: In radiography procedures, radiographers’ suboptimal positioning and exposure parameter settings may necessitate image retakes, subjecting patients to unnecessary ionizing radiation exposure. Reducing retakes is crucial to minimize patient X-ray exposure and conserve medical resources. OBJECTIVE: We propose a Digital Radiography (DR) Pre-imaging All-round Assistant (PIAA) that leverages Artificial Intelligence (AI) technology to enhance traditional DR. METHODS: PIAA consists of an RGB-Depth (RGB-D) multi-camera array, an embedded computing platform, and multiple software components. It features an Adaptive RGB-D Image Acquisition (ARDIA) module that automatically selects the appropriate RGB camera based on the distance between the cameras and patients. It includes a 2.5D Selective Skeletal Keypoints Estimation (2.5D-SSKE) module that fuses depth information with 2D keypoints to estimate the pose of target body parts. Thirdly, it also uses a Domain expertise (DE) embedded Full-body Exposure Parameter Estimation (DFEPE) module that combines 2.5D-SSKE and DE to accurately estimate parameters for full-body DR views. RESULTS: Optimizes DR workflow, significantly enhancing operational efficiency. The average time required for positioning patients and preparing exposure parameters was reduced from 73 seconds to 8 seconds. CONCLUSIONS: PIAA shows significant promise for extension to full-body examinations.

DSN-BR-Based Online Inspection Method and Application for Surface Defects of Pharmaceutical Products in Aluminum-Plastic Blister Packages

Ensuring high product quality is of paramount importance in pharmaceutical drug manufacturing, as it is subject to rigorous regulatory practices. This study presents a research focused on the development of an on-line detection method and system for identifying surface defects in pharmaceutical products packaged in aluminum-plastic blisters. Firstly, the aluminum-plastic blister packages exhibit multi-scale features and inter-class indistinction. To address this, the deep semantic network with boundary refinement (DSN-BR) model is proposed, which leverages semantic segmentation domain knowledge, to accurately segment the defects in pixel level. Additionally, a specialized image acquisition module that minimizes the impact of ambient light is established, ensuring high-quality image capture. Finally, the image acquisition module, image detection module, and data management module are designed to construct a comprehensive online surface defect detection system. To validate the effectiveness of our approach, we employ a real dataset for instance verification on the implemented system. The experimental results substantiate the outstanding performance of the DSN-BR, achieving the mean intersection over union (MIoU) of 90.5%. Furthermore, the proposed system achieves an inference speed of up to 14.12 f/s, while attaining an F1-Score of 98.25%. These results demonstrate that the system meets the actual needs of the enterprise and provides theoretical and methodological support for intelligent inspection of product surface quality. By standardizing the control process of pharmaceutical manufacturing and improving the management capability of the manufacturing process, our approach holds significant market application prospects.

A standalone and portable imaging detection system with embedded computing for automated defect inspection of microfluidic devices

The existing methods for defect detection in PDMS microfluidic chips typically involve complex image recognition algorithms or manual inspection and still lack efficiency and reliability. Although some automatic defect detection methods have been proposed in recent years, most of them still rely on external computation systems to deploy. To address these challenges, we propose an independent portable defect detection system with embedded computing for microfluidic devices. This portable system is completely self-contained, integrating an image acquisition module, a control panel module, a power module, and an embedded computing control module to realize chip detection, processing, and result display functions. Experimental results show that the system can effectively detect most of the commonly seen defects in PDMS-based microfluidic chips, proving to be more efficient and reliable than manual inspection. With the control of the embedded system, two detection methods: template matching (based on comparison with standard samples) and automatic defect detection (based on surface defect recognition) were used to identify defects in PDMS-based microfluidic chips. The proposed system can automatically inspect and analyze chips without the need for external laboratory support and can provide a promising solution for future microfluidic chip manufacturing and operation.

A mechanical arm based on image recognition and remote control

Mechanical arm, image recognition and wireless transmission have been a hot topic and technology in recent years. This paper mainly introduces the object recognition mechanical arm based on the mechanical arm and wireless transmission technology. The system is modular, and the whole equipment system is divided into remote control system and mechanical arm system according to function. The remote control system is an upper computer software which integrates image recognition, related data processing module and wireless transmission module. The remote end includes wireless transmission module, steering gear control module and image acquisition module. The image is collected and sent to the remote control system, and the motion trajectory of the mechanical arm is obtained after calculation, and the control signal of the stepping motor on the mechanical arm is further calculated, and finally sent to the mechanical arm system to achieve the purpose of remote control. At the same time, during the movement of the mechanical arm, the relevant data is collected by the sensor on the mechanical arm and sent back to the remote control terminal for display and judgment. In this project, the mechanical arm system can automatically recognize the target on a two-dimensional plane and move the mechanical arm according to the object position. Meanwhile, the upper computer can display relevant data information.

A computer vision approach to improve maintenance automation for thermal power plants lubrication systems

PurposeLubricating oil leakage is a common issue in thermal power plant operation sites, requiring prompt equipment maintenance. The real-time detection of leakage occurrences could avoid disruptive consequences caused by the lack of timely maintenance. Currently, inspection operations are mostly carried out manually, resulting in time-consuming processes prone to health and safety hazards. To overcome such issues, this paper proposes a machine vision-based inspection system aimed at automating the oil leakage detection for improving the maintenance procedures.Design/methodology/approachThe approach aims at developing a novel modular-structured automatic inspection system. The image acquisition module collects digital images along a predefined inspection path using a dual-light (i.e. ultraviolet and blue light) illumination system, deploying the fluorescence of the lubricating oil while suppressing unwanted background noise. The image processing module is designed to detect the oil leakage within the digital images minimizing detection errors. A case study is reported to validate the industrial suitability of the proposed inspection system.FindingsOn-site experimental results demonstrate the capabilities to complete the automatic inspection procedures of the tested industrial equipment by achieving an oil leakage detection accuracy up to 99.13%.Practical implicationsThe proposed inspection system can be adopted in industrial context to detect lubricant leakage ensuring the equipment and the operators safety.Originality/valueThe proposed inspection system adopts a computer vision approach, which deploys the combination of two separate sources of light, to boost the detection capabilities, enabling the application for a variety of particularly hard-to-inspect industrial contexts.

LEAF BLIGHT DETECTION USING DEEP LEARNING

Plant leaf blight is a common disease that affects many crops, causing significant economic losses in agriculture. Early detection of blight is essential for farmers to prevent the spread of the disease, control its severity, and improve crop yields. Manual detection of blight is time-consuming and often unreliable, making it necessary to use advanced technologies for accurate and efficient detection. Plant leaf blight detection systems can help farmers monitor their crops in real-time and identify blight at an early stage. This enables farmers to take appropriate measures such as crop rotation, timely application of fungicides, and improving plant nutrition to prevent the spread of the disease. By using plant leaf blight detection systems, farmers can reduce the economic losses caused by the disease and ensure the availability of healthy crops for consumption. This project presents a smart system for leaf blight detection that uses deep learning and convolutional neural networks (CNN) and is implemented on an Android platform. The system comprises an image acquisition module, a pre-processing module, and a leaf blight classification module. The image acquisition module captures images of leaves from a mobile camera, while the pre-processing module enhances the quality of images by removing noise and enhancing contrast. The leaf blight classification module uses a pre-trained CNN model to accurately classify images into healthy or diseased categories and provide the Confidence level of the Blight (disease) present.

Up-conversion Luminescence System for Quantitative Detection of IL-6.

Interleukin-6 (IL-6) is a very important cytokine and an early predictor of survival in febrile patients (eg, patients with COVID-19). With the outbreak of the COVID-19 in the world, the significance of medical detection of interleukin 6 has gradually become prominent. A method to point-of-care(POCT) diagnosis and monitoring of IL-6 levels in patients is urgently needed. In this work, an up-conversion luminescence system (ULS) based on upconverting nanoparticles (UCNs) for quantitative detection of IL-6 was designed. The ULS consists of Micro Controller Units (MCU), transmission device, laser, image acquisition module, Bluetooth module, etc. Through hardware system acquisition and image software algorithm processing, we obtain a limit of detection (LOD) of IL-6 at 1 ng/mL, and the quantitative range is from 1 to 200 ng/mL. The system is handheld and has great detection accuracy. The detection time is 10 minutes. In addition, the system can access mobile device terminals (smartphones, personal computers, etc.) or 5G cloud servers via Bluetooth and WIFI. Patients and family members can view medical data through mobile terminals, and the data stored in the 5G cloud server can be used for edge computing and big data analysis. It is suitable for the early diagnosis of infectious diseases such as COVID-19 and has good application prospects.

An Automated Image Processing Module for Quality Evaluation of Milled Rice.

The paper demonstrates a low-cost rice quality assessment system based on image processing and machine learning (ML) algorithms. A Raspberry-Pi based image acquisition module was developed to extract the structural and geometric features from 3081 images of eight different varieties of rice grains. Based on features such as perimeter, area, solidity, roundness, compactness, and shape factor, an automatic identification system is developed to segment the grains based on their types and classify them by using seven machine learning algorithms. These ML models are trained using the images and are compared using different ML models. ROC curves are plotted for each model for quantitative analysis to assess the model's performance. It is concluded that the random forest classifier presents an accuracy of 77 percent and is the best-performing model for the classification of rice varieties. Furthermore, the same algorithm is efficiently employed to determine the price of adulterated rice samples based upon the market price of individual rice.

Design and implementation of moving object detection based on Zynq

In recent years, video image processing technology has made rapid progress. Some algorithms related to moving target detection and tracking, relying on the rich resources of computers, make the implementation of these algorithms based on the PC platform has been relatively mature, but there are problems such as poor real-time performance, high power consumption, resource consumption. The Zynq hardware platform has both FPGA, which is good at parallel operation, and ARM core, which is good at control, which can effectively implement the algorithm transplantation of moving target detection and effectively improve the above shortcomings. The system takes Zynq hardware platform as the core unit, which is composed of video image acquisition module, image processing module, information transmission module, DDR cache module and display module. The PL side of Zynq implements video image acquisition, gray conversion, frame difference algorithm, expansion corrosion algorithm. The PS terminal mainly completes the configuration of IIC camera and VDMA. The output module finally realizes the output of HDMI through protocol conversion. This system has wide application value in moving target detection.

Custom-Developed Reflection-Transmission Integrated Vision System for Rapid Detection of Huanglongbing Based on the Features of Blotchy Mottled Texture and Starch Accumulation in Leaves.

Huanglongbing (HLB) is a highly contagious and devastating citrus disease that causes huge economic losses to the citrus industry. Because it cannot be cured, timely detection of the HLB infection status of plants and removal of diseased trees are effective ways to reduce losses. However, complex HLB symptoms, such as single HLB-symptomatic or zinc deficiency + HLB-positive, cannot be identified by a single reflection imaging method at present. In this study, a vision system with an integrated reflection-transmission image acquisition module, human-computer interaction module, and power supply module was developed for rapid HLB detection in the field. In reflection imaging mode, 660 nm polarized light was used as the illumination source to enhance the contrast of the HLB symptoms in the images based on the differences in the absorption of narrow-band light by the components within the leaves. In transmission imaging mode, polarization images were obtained in four directions, and the polarization angle images were calculated using the Stokes vector to detect the optical activity of starch. A step-by-step classification model with four steps was used for the identification of six classes of samples (healthy, HLB-symptomatic, zinc deficiency, zinc deficiency + HLB-positive, magnesium deficiency, and boron deficiency). The results showed that the model had an accuracy of 96.92% for the full category of samples and 98.08% for the identification of multiple types of HLB (HLB-symptomatic and zinc deficiency + HLB-positive). In addition, the classification model had good recognition of zinc deficiency and zinc deficiency + HLB-positive samples, at 92.86%.

From Patch to Pixel: A Transformer-Based Hierarchical Framework for Compressive Image Sensing

The convolutional neural network (CNN)-based reconstruction methods have dominated the compressive sensing (CS) in recent years. However, existing CNN-based approaches show potential restrictions in capturing non-local similarity of images, because of the intrinsic characteristic of convolutional layers, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathit{i.e.}$</tex-math></inline-formula> , locality and weight sharing. In parallel, the emerging Transformer architecture shows fine capacity in modeling long-distance correlations onto embedded tokens for language and images. Yet vanilla Transformer does not exceed CNN-based networks considerably but shows roughly comparable performance, and the culprit can be the missing of sophisticated inductive bias regarding the local image structures. In this article, to eliminate the restrictions of the aforementioned paradigms, we propose a Transformer-based hierarchical framework, dubbed TCS-Net, for compressive image sensing (or image compressive sensing) with a <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathit{patch-to-pixel}$</tex-math></inline-formula> manner. Concretely, the proposed TCS-Net consists of an image acquisition module and a reconstruction module (includes two key decoding phases: a patch-wise decoding phase and a pixel-wise decoding phase). The acquisition module can implement data-driven image sampling by jointly learning with the decoding phases. By adjusting the Transformer architecture to the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$\mathit{patch-to-pixel}$</tex-math></inline-formula> multi-stage pattern, our reconstruction module can gradually decode the CS measurements from the patch-wise outlines to the pixel-wise textures, thereby building a high-precision mapping for image reconstruction. Extensive experiments on several datasets verify that the proposed TCS-Net outperforms existing state-of-the-art image CS methods by considerable margins.

Compound minirhizotron device for root phenotype and water content near root zone

The study of phenotype plays a significant role in plant growth and development. In plant phenotype, the growth state and size distribution of roots are crucial to plant growth and development. Exploring the variation of root development and root zone water content is conducive to improving planting technology with efficiency. Whereas, commercial root phenotype monitoring devices on the market are expensive and sophisticated. This study aims to develop a compound minirhizotron device to acquire images of tomato root system and monitor water content near the root zone. The device mainly included core controller module, transmission module, image acquisition module and moisture monitoring module. The device could realize image acquisition of root system at multiple positions and monitoring of water content near root zone under the condition of less damage to root system. Through image processing, parameters such as root length and root diameter could be calculated. Results indicated that tomato roots at the depth of 6–10 cm from the substrate surface were more densely distributed, and root density decreased with increasing substrate depths. And furthermore, total root length showed a gradual downward trend with the downward movement of the substrate layer. In addition, the closer to the vertical downward straight line from the tomato cultivation site, the greater the root diameter within the monitoring range. In the monitoring results of water content near root zone, the closer to the ground, the higher monitored substrate water content. The overall results showed that the compound minirhizotron device could aquire both root system images and water content near root zone stablely and for a long-term, which could improve the efficiency of planting and irrigation technology.

Cooperative control method of robot formation movement path based on machine vision

Because the formation path information is not extracted in the process of robot formation motion path cooperative control, the image processing time is long and the effect of avoiding obstacles to reach the target position is poor. Therefore, a robot formation motion path cooperative control method based on machine vision is proposed. Through the design of embedded chip control module, image acquisition module and communication module, the application process of machine vision technology is designed, and the target features in the motion path of robot formation are extracted, including other robot targets and obstacle targets in the formation. Then, the formation path information is extracted based on image preprocessing. On this basis, using the formation form of virtual structure-driver following method, the obstacle avoidance control of robot formation motion path is completed by potential field method. The experimental results show that after using this method, the images of its CCD camera when performing various processing are less than 200 ms. The moving environment image shows that this method has high processing performance for robot formation, and under the control of this method, the formation can change the formation and intelligent robot to avoid obstacles and reach the target position.

Design of field real-time target spraying system based on improved YOLOv5.

Deep learning techniques have made great progress in the field of target detection in recent years, making it possible to accurately identify plants in complex environments in agricultural fields. This project combines deep learning algorithms with spraying technology to design a machine vision precision real-time targeting spraying system for field scenarios. Firstly, the overall structure scheme of the system consisting of image acquisition and recognition module, electronically controlled spray module and pressure-stabilized pesticide supply module was proposed. After that, based on the target detection model YOLOv5s, the model is lightened and improved by replacing the backbone network and adding an attention mechanism. Based on this, a grille decision control algorithm for solenoid valve group on-off was designed, while common malignant weeds were selected as objects to produce data sets and complete model training. Finally, the deployment of the hardware system and detection model on the electric spray bar sprayer was completed, and field trials were conducted at different speeds. The experimental results show that the improved algorithm reduces the model size to 53.57% of the original model with less impact on mAP accuracy, improves FPS by 18.16%. The accuracy of on-target spraying at 2km/h, 3km/h and 4km/h speeds were 90.80%, 86.20% and 79.61%, respectively, and the spraying hit rate decreased as the operating speed increased. Among the hit rate components, the effective recognition rate was significantly affected by speed, while the relative recognition hit rate was less affected.

Robot Structural Optimization Based on Computer Intelligent Image

In order to solve the problems of low degree of automation, difficult identification of picking objects, and large picking damage in traditional fruit and vegetable picking operations, the author proposes a robot structure optimization method based on computer intelligent images. This method introduces the computer imaging technology, examines the principle of imaging the computer imaging technology as the in-depth study of the computer imaging technology, completes the mechanical design of the selected robot, and optimizes all hardware models of choice robot. At the same time, the computer image acquisition system, image acquisition module, and execution module are designed; finally, the computer image information processing flow design of the picking robot is completed, and the simulation experiment of the picking robot is carried out. Experimental results show that in the experiments with 166, 142, and 165 tomato identification numbers, the identification accuracy rates were all over 96%. Conclusion. The picking robot based on computer images has a simple structure, high recognition accuracy of picking targets, less damage to the picking targets, high safety and stability, and great promotion value.

Design and Test of the Seedling Cavitation and Lodging Monitoring System for the Rape Blanket Seedling Transplanter

To realize the real-time monitoring of the lodging and cavitation state of the rape blanket seedling transplanter, a real-time online monitoring system set for cavitation and lodging based on machine vision was designed. The system was mainly composed of an image acquisition module, a height sensor, and an IPC (Industrial Personal Computer). When the system starts to work, the height of the lens from the ground and the collected images are saved in real time, and the operating status of the equipment is judged according to the height of the lens from the ground, and the actual corresponding size of the image pixels is corrected. We proposed a calculation method for judging the cavitation and lodging state of the seedlings based on the comparison between the actual planting position and the theoretical planting position of the seedlings. The processing and analysis software for the cavitation and lodging monitoring system of the rape blanket transplanter was compiled, and the monitoring system was developed. The dynamic test showed that the maximum absolute error of the cavitation recognition rate was 1.23%, and the maximum absolute error of the lodging recognition rate was 3.80%. The total time consumed of a 1 m image collected in real time from the collection, image processing and stitching, to seedling status recognition of the rape blanket seedling field was about 1 s. The accuracy and system processing time of the seedling cavitation and lodging state judgment can meet the accuracy and real-time requirements of cavitation and lodging state recognition for field transfer of rape blanket seedlings.

Research on Methods of Physical Aided Education Based on Deep Learning

In order to better meet the training needs of sports and improve the standardization of sports training, an openpose-based sports posture estimation method and assisted training system are proposed, combining the basic structure and principle of openpose network. Firstly, the human posture estimation algorithm is constructed by combining with the openpose network; secondly, the overall framework, specific operation process, image acquisition, posture estimation, and other modules of the sports assistance system are designed in detail; finally, the openpose posture estimation method constructed above is validated. The results show that the value of the loss function obtained by the algorithm gradually stabilizes after 250 iterations. By using the COCO dataset as the training base and comparing it with the standard posture, it is found that the algorithm can correctly identify different badminton action postures, and the recognition rate can reach up to 94%. This shows that the algorithm is feasible and can be used for posture estimation and training of badminton sports movements.

A Measuring Device for Residual Thickness of Brake Shoe of Railcar Based on Deep Learning

At present, the residual thickness of brake shoe of metro vehicles is mainly measured manually by naked eyes or simple ruler, which has some problems such as low detection efficiency, large subjective randomness and poor engineering applicability. Therefore, this paper proposes a device for measuring the residual thickness of brake shoes of railcar based on deep learning method. Firstly, a residual thickness measuring device includes parts of image acquisition module, image positioning and segmentation module, and image correction module. Then for each module, the article conducted the related experiments for verifying the accuracy of the system. The results show that the brake shoe residual thickness measurement device proposed in this paper has higher accuracy and better robustness than manual detection, and has obvious advantages in detection time consumption.

Artificial Intelligence in the Auxiliary Guidance Function of Athletes’ Movement Standard Training in Physical Education

Artificial Intelligence (AI) is a new technical science that studies and develops theories, methods, techniques and application systems for simulating, extending and expanding human intelligence. This paper is based on the research of AI in the auxiliary guidance function of athletes’ standard training in physical education. It aims to conduct data mining from different aspects such as different joint angular speeds, motion injury screening and different parts of sports injuries and then integrate these aspects. Create a way to reduce athletes’ injuries and scientific training. In order to improve the recognition efficiency of athletes’ movement patterns, nonlinear auto regressive neural networks are used to recognize the movement patterns of athletes’ limb surface signals. Through this research work, it can provide a certain reference basis and practice platform for the research on the auxiliary guidance role of AI in the sports standard training of athletes in physical education. Performance design and implementation include four modules: Image acquisition, preprocessing, motion detection and human motion recognition. Between them, the image acquisition module uses a memory mapping path to determine the format of the camera frame image, and the image format conversion is completed through channel conversion. Experimental data shows that athletes’ strength quality, speed of action response, technical continuity, psychological stability and physical control ability have all been greatly improved. Among them, the most obvious is that with the assistance of AI technology, the psychological stability has reached 9.2; the strength quality has reached 9.1.