Template Matching Algorithm Research Articles

Remote Sensing Multimodal Image Matching Based on Structure Feature and Learnable Matching Network

Matching remotely sensed multimodal images is a crucial process that poses significant challenges due to nonlinear radiometric differences and substantial image noise. To overcome these difficulties, this study presents a novel and practical template-matching algorithm specifically designed for this purpose. Unlike traditional approaches that rely on image intensity, the proposed algorithm focuses on matching multimodal images based on their geometric structure information. This approach enables the method to effectively adapt to variations in grayscale caused by radiometric differences. To enhance the matching performance, principal component analysis calculation based on the log-Gabor filter is proposed to estimate the structural feature of the image. The proposed method can estimate the structure feature accurately even under severe noise distortion. In addition, a learnable matching network is proposed for similarity measuring to adapt to the gradient reversal caused by the radiometric difference among remotely sensed multimodal images. Infrared, visible light, and synthetic aperture radar images are adopted for the evaluation, to verify the performance of the proposed algorithm. Based on the results, the proposed algorithm has a distinct advantage over other state-of-the-art template-matching algorithms.

Fast template matching and object detection techniques using φ-correlation and binary circuits

In this paper, two new accelerated and robust template matching and object detection algorithms are established. The algorithms retain the accuracy measures compared to relevant well-known standard and recent schemes, and remarkably accelerate the execution time simultaneously. Instead of dealing with images in the grayscale or color representations, the proposed methods are established by measuring resemblance on extracted binary images. The two different techniques are based on the φ-correlation coefficient and logical circuits. Both techniques enhance the accuracy, compared to two classical and two recent methods, while giving very efficient running time. The algorithms are carried out on two different datasets used in state-of-the-art methods with performance and comparison analysis. Robustness of the proposed algorithms against normal and artificial noises is examined and assured as well.

Improving the accuracy and speed of fast template‐matching algorithms by neural architecture search

AbstractNeural architecture search can be used to find convolutional neural architectures that are precise and robust while enjoying enough speed for industrial image processing applications. In this paper, our goal is to achieve optimal convolutional neural networks (CNNs) for multiple‐templates matching for applications such as licence plates detection (LPD). We perform an iterative local neural architecture search for the models with minimum validation error as well as low computational cost from our search space of about 32 billion models. We describe the findings of the experience and discuss the specifications of the final optimal architectures. About 20‐times error reduction and 6‐times computational complexity reduction is achieved over our engineered neural architecture after about 500 neural architecture evaluation (in about 10 h). The typical speed of our final model is comparable to classic template matching algorithms while performing more robust and multiple‐template matching with different scales.

Small target location algorithm based on optimized template matching

At present, few published template matching algorithms can achieve the performance of commercial software. This paper studies and optimizes the existing geometric feature template matching algorithm, and combines it with the improved Fourier Merlin transform registration algorithm, and finally obtains a fast small target location algorithm. The matching time of the algorithm is 1-4ms, its positioning accuracy is sub-pixel level, the angle accuracy is less than 1 °, and it has good robustness to illumination. The experimental data show that the algorithm meets the requirements of real-time and accurate industrial automation production.

Fast Localization and High Accuracy Recognition of Tire Surface Embossed Characters Based on CNN

To solve the problem of recognizing artificial tire-side pressure printing characters with low efficiency and high labor intensity, we propose a CNN-based method for tire surface character recognition. In the image pre-processing, the SSR algorithm is improved to enhance the contrast of characters, and the Normalized Cross Correlation template matching algorithm based on pyramid acceleration is proposed to quickly locate the “DOT” characters and segment them. The improved LeNet-5 network structure is used to recognize characters, and a self-built digital sample library is randomly divided according to the ratio of 8:2 to conduct digital recognition experiments. The experimental results show that the recognition accuracy of the training set can reach 95.9%, and the accuracy of the validation set is 99.5%. The accuracy of the testing set is 95.6%, which meets the practical application requirements. Moreover, the whole algorithm only needs to be implemented on a commonly configured CPU, reducing equipment costs.

Progressive Classifier Mechanism for Bridge Expansion Joint Health Status Monitoring System Based on Acoustic Sensors.

The application of IoT (Internet of Things) technology to the health monitoring of expansion joints is of great importance in enhancing the efficiency of bridge expansion joint maintenance. In this study, a low-power, high-efficiency, end-to-cloud coordinated monitoring system analyzes acoustic signals to identify faults in bridge expansion joints. To address the issue of scarce authentic data related to bridge expansion joint failures, an expansion joint damage simulation data collection platform is established for well-annotated datasets. Based on this, a progressive two-level classifier mechanism is proposed, combining template matching based on AMPD (Automatic Peak Detection) and deep learning algorithms based on VMD (Variational Mode Decomposition), denoising, and utilizing edge and cloud computing power efficiently. The simulation-based datasets were used to test the two-level algorithm, with the first-level edge-end template matching algorithm achieving fault detection rates of 93.3% and the second-level cloud-based deep learning algorithm achieving classification accuracy of 98.4%. The proposed system in this paper has demonstrated efficient performance in monitoring the health of expansion joints, according to the aforementioned results.

Research on Identification and Location of Charging Ports of Multiple Electric Vehicles Based on SFLDLC-CBAM-YOLOV7-Tinp-CTMA

With the gradual maturity of autonomous driving and automatic parking technology, electric vehicle charging is moving towards automation. The charging port (CP) location is an important basis for realizing automatic charging. Existing CP identification algorithms are only suitable for a single vehicle model with poor universality. Therefore, this paper proposes a set of methods that can identify the CPs of various vehicle types. The recognition process is divided into a rough positioning stage (RPS) and a precise positioning stage (PPS). In this study, the data sets corresponding to four types of vehicle CPs under different environments are established. In the RPS, the characteristic information of the CP is obtained based on the combination of convolutional block attention module (CBAM) and YOLOV7-tinp, and its position information is calculated using the similar projection relationship. For the PPS, this paper proposes a data enhancement method based on similar feature location to determine the label category (SFLDLC). The CBAM-YOLOV7-tinp is used to identify the feature location information, and the cluster template matching algorithm (CTMA) is used to obtain the accurate feature location and tag type, and the EPnP algorithm is used to calculate the location and posture (LP) information. The results of the LP solution are used to provide the position coordinates of the CP relative to the robot base. Finally, the AUBO-i10 robot is used to complete the experimental test. The corresponding results show that the average positioning errors (x, y, z, rx, ry, and rz) of the CP are 0.64 mm, 0.88 mm, 1.24 mm, 1.19 degrees, 1.00 degrees, and 0.57 degrees, respectively, and the integrated insertion success rate is 94.25%. Therefore, the algorithm proposed in this paper can efficiently and accurately identify and locate various types of CP and meet the actual plugging requirements.

A new machine vision detection method for identifying and screening out various large foreign objects on coal belt conveyor lines

Large foreign object transporting by coal mine conveyor belt may lead to production safety hazards. To reduce safety accidents during coal mining, a large foreign object detection method based on machine vision is proposed in this paper. An adaptive weighted multi-scale Retinex (MSR) image enhancement algorithm is proposed to improve the captured image quality of the belt conveyor line. An improved multi-scale template matching algorithm is designed by combining the frame difference and area methods to screen and identify large foreign objects mixed in coals. The multi-layer perceptron (MLP) network optimized by the Gray Wolf algorithm is introduced to identify the large objects. Experimental results show that the identification accuracy reaches 98.8% for the large foreign objects. Furthermore, industrial field test is carried out in the Gaoyang coal mine, and the filed test results demonstrate that the identification accuracy of the proposed method is more than 95%. Hence, the proposed method meets the industrial detection requirements and can be used in practices for detecting the large foreign objects.

SIFNet: A self-attention interaction fusion network for multisource satellite imagery template matching

Multisource satellite images provide abundant and complementary earth observations, while nonlinear radiometric and geometric distortions (such as scale and rotation variations) between these multimodal images pose remarkable challenges for further remote sensing applications, such as change detection. We therefore proposed a template matching algorithm based on self-attention interactive fusion network, named SIFNet, to align multisource satellite images. First, a feature pyramid network was first conducted to extract multiscale features for each pixel, with the template and reference images as the inputs. Then, the extracted features were fused by self-attention layers in Transformer for information interaction. Third, the similarity and semantic matching loss functions were developed to convert satellite imagery registration into regression task, allowing SIFNet aligning multimodal patch images more efficiently based on point-to-point correspondence, instead of globally searching extremums as previous matching strategies did. We performed experiments based on four multimodal datasets (i.e. Google, GF-2, Landsat-8 and optical-SAR images) with various scenes to evaluate the performance and robustness of SIFNet. The results demonstrate the proposed SIFNet performed a comparable accuracy for template matching with other algorithms and was robust to geometric distortions and radiometric variations of multisource remote sensing data.

3d imaging and analysis of cracks in loaded concrete samples

Concrete plays a central role as the standard building material in civil engineering. Experimental characterization of the concrete microstructure and a description of failure mechanisms are important to understand the concrete’s mechanical properties. Computed tomography is a powerful source of information as it yields 3d images of concrete specimens. However, complete visual inspection is often infeasible due to very large image sizes. Hence, automatic methods for crack detection and segmentation are needed. A region-growing algorithm and a 3d U-Net showed promising results in a previous study. Cracks in normal concrete and high-performance concrete that were initiated via tensile tests were investigated. Here, the methods are validated on a more diverse set of concrete types and crack characteristics. Adequate adaptions of the methods are necessary to deal with the complex crack structures. The segmentation results are assessed qualitatively and compared to those of a template matching algorithm which is well-established in industry.

Transfer learning for real-time crater detection on asteroids using a Fully Convolutional Neural Network

This paper proposes the use of Transfer Learning from the Moon to Ceres for real-time autonomous crater detection operations on asteroids. This approach, based on the use of a U-Net for crater segmentation, starts with the network training on a Moon Digital Elevation Model; the trained model is then refined using two different strategies, one based on data augmentation and one based on fine tuning with small sets of Ceres images. Data from the DAWN mission was used, while the reference crater catalogue for Ceres is the one created by Zeilnhofer from LAMO data, containing 44,594 craters with diameter greater than 1 km in a latitude range of 84.66°S–89.62°N, full longitude. The loss function chosen during network training is the Focal Tversky Loss, which takes into consideration the imbalanced distribution of class pixels in the ground truth (pixels corresponding to crater rims represent approximately only 3% over the whole dataset). For both the Moon and Ceres, the U-Net output is post-processed with a template matching algorithm. The detected craters are counted and compared to the crater catalogue locations in order to compute metrics performances (precision, recall, F1 score) in the crater domain. Post-processing on the Moon reached performance of 84.12% F1 score on the baseline model. The baseline and augmented models gave a lower performance on Ceres data, although comparable with the outcomes found in other works, with a best F1 score of 70.17%. However, when the model is completely retrained starting from the baseline, it ends up with a best F1 score of 79.96%. Finally, a strategy for the actual on-board implementation is proposed considering the NVIDIA Jetson TX2 accelerator. We demonstrated that the inference time of the model on the accelerator is on the order of few milliseconds without loss of performances, thus proving the possibility to consider this approach for future real-time on-board navigation systems.

Micro-amplitude vibration measurement using vision-based magnification and tracking

A new deep learning-based vision measurement method is proposed to accurately measure the micro-vibration displacements of objects in different illuminations and backgrounds. The measurement method pre-process the video, then the deep learning correlation methods are used to zoom in the target object and track the vibration trajectory, and the pixel displacement is converted to actual displacement by pixel equivalents. By comparing the three sets of experiments, the proposed method has exceptional accuracy. When measuring vibration displacement of 0.1 mm, the Root Mean Square Error (RMSE) and Normalized Root Mean Square Error (NRMSE) are 0.0234 mm and 11.8601 %. By comparing with the Rectangle Detection Algorithm and the Template Matching Algorithm, the proposed algorithm outperforms these two traditional methods, especially for the complex environments. It can be concluded that this method, as a new visual measurement method, can be adapted to a variety of complex environments and can accurately measure micro-amplitude vibrations.

Soccer Robot Localization Detection Model Based on Monte Carlo Particle Filter and Template Matching Algorithm

Soccer Robot Localization Detection Model Based on Monte Carlo Particle Filter and Template Matching Algorithm

3D Feature-Based Sampled-Data Visual Tracking*

3D feature-based Visual Servoing (VS) on the one hand shows attractive peculiarities, on the other hand it suffers from drawbacks related to the existence of local minima, which may affect the convergence character of the VS control loop. Furthermore, the performance of the visual tracking module may constitute a bottleneck enforcing severe constraints on the workspace and visual task execution speed. In this paper we introduce a novel sampled-data model of the 3D feature-based VS, and, in order to avoid drawbacks due to local minima, we plan the target reference trajectory in the feature space with the aim to constraint the feature error dynamics to remain close to the desired equilibrium point. Then, we propose a novel feature generation based on the homography provided by a template matching algorithm based on the Zero mean Normalized Cross Correlation (ZNCC) and the design of a visual tracking scheme by resorting to the Extended Kalman Filter (EKF) and Lyapunov direct method, which explicitly takes into account the camera velocity limits, while ensuring stability.

Template matching via bipartite graph and graph attention mechanism

AbstractMost existing template matching algorithms are global matching between the template target and the search region, which makes the matching process retain a lot of unfavourable background information and ignore the structure and local information of the template target. To address this problem, a template matching algorithm based on bipartite graph and graph attention mechanism is proposed in this paper. The algorithm models the similarity matching problem between template features and search region features as a complete bipartite graph, realises local‐to‐local information transfer between the two, and uses the graph attention mechanism to apply weights between local information to obtain a learnable embedding network module. In addition, in terms of feature representation, a multi‐level feature fusion module based on CNN is introduced, which improves the representation of a target by fusing features with different representational meanings of the target. Experimental results on several typical datasets show that the proposed algorithm achieves leading performance in terms of accuracy and efficiency compared to the two state‐of‐the‐art CNN‐based template matching algorithms, Deep‐DIM and QATM.

An Automatic Detection Method for Cutting Path of Chips in Wafer.

Microscopic imaging is easily affected by the strength of illumination, and the chip surface qualities of different wafers are different. Therefore, wafer images have defects such as uneven brightness distribution, obvious differences in chip region characteristics, etc., which affect the positioning accuracy of the wafer cutting path. For this reason, this thesis proposes an automatic chip-cutting path-planning method in the wafer image of the Glass Passivation Parts (GPPs) process without a mark. First, the wafer image is calibrated for brightness. Then, the template matching algorithm is used to determine the chip region and the center of gravity position of the chip region. We find the position of the geometric feature (interlayer) in the chip region, and the interlayer is used as an auxiliary location to determine the final cutting path. The experiment shows that the image quality can be improved, and chip region features can be highlighted when preprocessing the image with brightness calibration. The results show that the average deviation of the gravity coordinates of the chip region in the x direction is 2.82 pixels. We proceeded by finding the interlayer in the chip region, marking it with discrete points, and using the improved Random Sample Consensus (RANSAC) algorithm to remove the abnormal discrete points and fit the remaining discrete points. The average fitting error is 0.8 pixels, which is better than the least squares method (LSM). The cutting path location algorithm proposed in this paper can adapt to environmental brightness changes and different qualities of chips, accurately and quickly determine the cutting path, and improve the chip cutting yield.

Design and implementation of Visual Epigynous Machine on Robotic Sorting

Robotic assembly is an important process to replace manual picking in industrial smart manufacturing. One of the important steps in assembly is robotics grasping and sorting of the workpiece. In this paper, a lightweight visual epigynous system was designed to address sorting workpieces with high efficiency, portability, and accuracy. The software of the system was divided into a vision calculation module (VCM), an ethernet communication module, a system control module, and a real-time display module. VCM is responsible for obtaining the information of workpieces, which includes segmenting the images captured by the webcam sequentially and recognizing the digit, and calculating the angle by algorithms of AKAZE feathers matching with and orientation estimation for vector comparison. The proposed method in VCM performs best in estimating the orientation of the workpiece through the comparison of different template matching algorithms. The industrial field experiments indicated that the robot could satisfy the requirement of visual sorting. This system provides a practical and robust solution for industrial workpiece sorting applications.

RETRACTED: Theoretical research on rice and wheat lodging detection based on artificial intelligence technology and a template matching algorithm

AbstractTraditional detection of lodging in rice (Oryza sativa L.) and wheat (Triticum aestivum L.) mostly uses statistical methods or routine ground surveys. These methods have a low degree of automation and poor accuracy. In order to improve the detection of rice and wheat lodging, this study combined artificial intelligence technology and a template matching algorithm to construct a rice and wheat lodging detection system. Moreover, this study used perceptual hash technology for image matching processing, built an intelligent recognition model on the basis of improved algorithms, and combined this with camera technology to build the system's functional framework. The constructed system could perform image matching for rice and wheat, and detected rice and wheat lodging on the basis of image matching. Finally, this study analyzed the system's function structure based on the actual situation of rice and wheat lodging, and designed experiments to verify the effect of the system. The experimental research results showed that the system constructed in this study can play an effective role in the detection of rice and wheat lodging.

Fast template matching in multi-modal image under pixel distribution mapping

A scheme that can achieve fast template matching by both traditional methods and neural network methods, called Hybrid Matching by Pixel Distribution Mapping (HMPDM), is proposed for the matching difficulty caused by the problem of nonlinear intensity differences between multi-modal images. The HMPDM scheme first extracts the distribution information of image pixels through the developed expanded Slice Transform (eSLT) matrix, to overcome the nonlinear intensity differences between visible and thermal infrared images; and then maps the eSLT matrices corresponding to different modal images into correlation surface images with the same modality through traditional integer mapping mechanism or neural network mapping mechanism; finally, template matching between visible and thermal infrared images is achieved by evaluating the similarity of correlation surface images through the Zero-mean Normalized Cross-Correlation (ZNCC) algorithm. The fast and classic ZNCC algorithm ensures the real-time performance of the HMPDM scheme in the most time-consuming sliding window matching stage. The traditional integer mapping mechanism and the neural network mapping mechanism enable the HMPDM scheme to use the traditional method and the neural network method alone to achieve template matching, or to combine the two methods to achieve template matching. The comprehensive utilization of traditional methods and neural network methods makes the HMPDM scheme adaptable to various platforms with limited hardware resources. The experimental results show that, on a hardware platform with limited resources, it only takes 0.18 s for a 64 × 64 template image to slide on a 256 × 256 query image to achieve template matching; and it only takes at least 1.2 s for neural network training on the CPU. The matching performance is also better than many popular multi-modal image template matching algorithms.

Toward Robust Partial-Image Based Template Matching Techniques for MRI-Guided Interventions.

We have developed an MRI-safe needle guidance toolkit for MRI-guided interventions intended to enable accurate positioning for needle-based procedures. The toolkit allows intuitive and accurate needle angulation and entry point positioning according to an MRI-based plan, using a flexible, patterned silicone 2D grid. The toolkit automatically matches the grid on MRI planning images with a physical silicon grid placed conformally on the patient's skin and provides the Interventional Radiologist an easy-to-use guide showing the needle entry point on the silicon grid as well as needle angle information. The radiologist can use this guide along with a 2-degree-of-freedom (rotation and angulation relative to the entry point) hand-held needle guide to place the needle into the anatomy of interest. The initial application that we are considering for this toolkit is arthrography, a diagnostic procedure to evaluate the joint space condition. However, this toolkit could be used for any needle-based and percutaneous procedures such as MRI-guided biopsy and facet joint injection. For matching the images, we adopt a transformation parameter estimation technique using the phase-only correlation method in the frequency domain. We investigated the robustness of this method against rotation, displacement, and Rician noise. The algorithm was able to successfully match all the dataset images. We also investigated the accuracy of identifying the entry point from registered template images as a prerequisite for a future targeting study. Application of the template matching algorithm to locate the needle entry points within the MRI dataset resulted in an average entry point location estimation accuracy of 0.12 ±0.2 mm. This promising result motivates a more detailed assessment of this algorithm in the future including a targeting study on a silicon phantom with embedded plastic targets to investigate the end-to-end accuracy of this automatic template matching algorithm in the interventional MRI room.