Difference Of Gaussian Research Articles

Blind image quality assessment using Beltrami filter-based contrast features (BF-bCF) & LSTM network

ABSTRACT Advanced networks excel in Blind Image Quality Assessment (BIQA), however accurate estimation of quality scores remains challenging due to unrevealed features extricated under various distortions. This article presents a Beltrami filter-based Contrast feature and long short-term memory (BF-bCF & LSTM) framework for feature extraction, suitable for both authentic and synthetic image distortions. The framework comprises four modules: an edge preservation and enhancement module, a distortion-conscious module, a weight module, and a quality-predicting network. The edge enhancement module assists the distortion-conscious module by enhancing the overall quality of the input grayscale image. The distortion-conscious module uses a modified Difference of Gaussian (DoG) measure to mitigate the distortion and assign patch weights in the filtered image. The proposed BIQA framework assimilating LSTM effectively extricates significant components using principal component analysis (PCA) and predicts nearer scores. Experiments on synthetic and authentic datasets showed superior performance relating to SROCC and PLCC above 0.9.

Accommodative and Vergence Responses to a Moving Stimulus in Concussion.

Concussed adolescents often report visual symptoms, especially for moving targets, but the mechanisms resulting in oculomotor deficits remain unclear. We objectively measured accommodative and vergence responses to a moving target in concussed adolescents and controls. Thirty-two symptomatic concussed participants (mean age, 14.4 ± 2.6years; mean days since concussion, 107days; range, 36-273days) and 32 healthy controls (mean age, 12.7 ± 2.1years) viewed a movie binocularly (closed-loop) and monocularly (vergence open-loop), as well as a Difference of Gaussians (DoG) target binocularly (accommodation open-loop). The movie or DoG target sinusoidally moved toward and away from participants at a 0.1-hertz (Hz) frequency at four separate stimulus amplitudes (1.50 diopters [D], 1.00 D, 0.50 D, 0.25 D) around a 2.50-D midpoint. Accommodation and vergence were continuously measured at 50 Hz using the PowerRef 3. Fourier analysis was used to assess the response amplitudes at the 0.1-Hz frequency. A 2 × 3 analysis of variance with the factors group (concussed, control) and viewing condition (binocular, monocular, DoG) was conducted on response amplitudes. Across groups, accommodative and vergence responses were significantly higher in binocular than monocular conditions (P < 0.001), but not DoG conditions. Compared to controls, concussed participants had significantly reduced monocular accommodative responses (P < 0.012; e.g., at 1.50 D, controls = 1.09 ± 0.47 D and concussed = 0.80 ± 0.36 D, P = 0.011). No group differences were observed for vergence responses in any viewing condition. Accommodative and vergence responses to the moving target were largely driven by disparity cues for both groups, with only minimal improvements in the presence of additional blur cues. Concussed participants showed reduced accommodative responses to a 0.1-Hz stimulus in monocular conditions, indicating mild accommodative deficits in the absence of disparity cues.

Diabetic retinopathy data augmentation and vessel segmentation through deep learning based three fully convolution neural networks

ProblemThe eye fundus imaging is used for early diagnosis of most damaging concerns such as diabetic retinopathy, retinal detachments and vascular occlusions. However, the presence of noise, low contrast between background and vasculature during imaging, and vessel morphology lead to uncertain vessel segmentation. AimThis paper proposes a novel retinalblood vessel segmentation method for fundus imaging using a Difference of Gaussian (DoG) filter and an ensemble of three fully convolutional neural network (FCNN) models. MethodsA Gaussian filter with standard deviation σ1 is applied on the preprocessed grayscale fundus image and is subtracted from a similarly applied Gaussian filter with standard deviation σ2 on the same image. The resultant image is then fed into each of the three fully convolutional neural networks as the input. The FCNN models' output is then passed through a voting classifier, and a final segmented vessel structure is obtained.The Difference of Gaussian filter played an essential part in removing the high frequency details (noise) and thus finely extracted the blood vessels from the retinal fundus with underlying artifacts. ResultsThe total dataset consists of 3832 augmented images transformed from 479 fundus images. The result shows that the proposed method has performed extremely well by achieving an accuracy of 96.50%, 97.69%, and 95.78% on DRIVE, CHASE,and real-time clinical datasets respectively. ConclusionThe FCNN ensemble model has demonstrated efficacy in precisely detecting retinal vessels and in the presence of various pathologies and vasculatures.

DEDBNet: DoG-enhanced dual-branch object detection network for remote sensing object detection

With the improvement of spatial resolution of remote sensing images, object detection of remote sensing images has gradually become a difficult task. Extracted object features are usually hidden in a large amount of interference information in the background due to the complexity and large area of backgrounds, as well as the multi-scale nature of objects in remote sensing images. Still, many existing background weakening methods face difficulties in practical applications and are prone to high rates of false positives and false negatives. Therefore, remote sensing object detection has become increasingly challenging. To address these challenges, a novel background weakening method called Difference of Gaussian (DoG) to weaken background (DWB) module is proposed. Then, we develop a dual-branch network, named DoG-Enhanced Dual-Branch Object Detection Network (DEDBNet) for Remote Sensing Object Detection. The base branch network is responsible for detecting objects, while the DWB's branch network corrects the detected objects using feature-level attention. To combine the features of these branches, we propose two new methods Self-Mutual-Correcter with Detect heads (SMCD) for corrective learning and Map Channel Attention (MCA) for channel attention. Self-Corrector (SC) enables modification and integration of features, while the Mutual-Corrector (MC) enhances the features and further fuses them. We evaluate our proposed network, DEDBNet, through extensive experiments on four public datasets (DOTA with an mAP of 0.836, DIOR with an mAP of 0.871, NWPU VHR-10 with an mAP of 0.973, and RSOD with an mAP of 0.975). The results demonstrate that our method outperforms other state-of-the-art object detection methods significantly for remote sensing images.

Significant effect of image contrast enhancement on weld defect detection

Weld defect inspection is an essential aspect of testing in industries field. From a human viewpoint, a manual inspection can make appropriate justification more difficult and lead to incorrect identification during weld defect detection. Weld defect inspection uses X-radiography testing, which is now mostly outdated. Recently, numerous researchers have utilized X-radiography digital images to inspect the defect. As a result, for error-free inspection, an autonomous weld detection and classification system are required. One of the most difficult issues in the field of image processing, particularly for enhancing image quality, is the issue of contrast variation and luminosity. Enhancement is carried out by adjusting the brightness of the dark or bright intensity to boost segmentation performance and image quality. To equalize contrast variation and luminosity, many different approaches have recently been put forth. In this research, a novel approach called Hybrid Statistical Enhancement (HSE), which is based on a direct strategy using statistical data, is proposed. The HSE method divided each pixel into three groups, the foreground, border, and problematic region, using the mean and standard deviation of a global and local neighborhood (luminosity and contrast). To illustrate the impact of the HSE method on the segmentation or detection stage, the datasets, specifically the weld defect image, were used. Bernsen and Otsu’s methods are the two segmentation techniques utilized. The findings from the objective and visual elements demonstrated that the HSE approach might automatically improve segmentation output while effectively enhancing contrast variation and normalizing luminosity. In comparison to the Homomorphic Filter (HF) and Difference of Gaussian (DoG) approaches, the segmentation results for HSE images had the lowest result according to Misclassification Error (ME). After being applied to the HSE images during the segmentation stage, every quantitative result showed an increase. For example, accuracy increased from 64.171 to 84.964. In summary, the application of the HSE method has resulted in an effective and efficient outcome for background correction as well as improving the quality of images.

Enhancing Focus Volume through Perceptual Focus Factor in Shape-from-Focus

Shape From Focus (SFF) reconstructs a scene’s shape using a series of images with varied focus settings. However, the effectiveness of SFF largely depends on the Focus Measure (FM) used, which is prone to noise-induced inaccuracies in focus values. To address these issues, we introduce a perception-influenced factor to refine the traditional Focus Volume (FV) derived from a traditional FM. Owing to the strong relationship between the Difference of Gaussians (DoG) and how the visual system perceives edges in a scene, we apply it to local areas of the image sequence by segmenting the image sequence into non-overlapping blocks. This process yields a new metric, the Perceptual Focus Factor (PFF), which we combine with the traditional FV to obtain an enhanced FV and, ultimately, an enhanced depth map. Intensive experiments are conducted by using fourteen synthetic and six real-world data sets. The performance of the proposed method is evaluated using quantitative measures, such as Root Mean Square Error (RMSE) and correlation. For fourteen synthetic data sets, the average RMSE measure of 6.88 and correction measure of 0.65 are obtained, which are improved through PFF from an RMSE of 7.44 and correlation of 0.56, respectively. Experimental results and comparative analysis demonstrate that the proposed approach outperforms the traditional state-of-the-art FMs in extracting depth maps.

Biomaterial-Based Biomimetic Visual Sensors: Inkjet Patterning of Bacteriorhodopsin.

Biomimetic visual sensors utilizing bacteriorhodopsin (bR) were fabricated by using an inkjet method. The inkjet printer facilitated the jetting of the bR suspension, allowing for the deposition of bR films. The resulting inkjet-printed bR film exhibited time-differential photocurrent response characteristics similar to those of a dip-coated bR film. By adjusting the number of printed bR film layers, the intensity of the photocurrent could be easily controlled. Moreover, the inkjet printing technique enabled unconstrained patterning, facilitating the design of various visual information processing functions, such as visual filters. In this study, we successfully fabricated two visual filters, namely, a two-dimensional Difference of Gaussian (DOG) filter and a Gabor filter. The printed DOG filter demonstrated edge detection capabilities corresponding to contour recognition in visual receptive fields. On the other hand, the printed Gabor filter proved effective in detecting objects of specific sizes as well as their motion and orientation. The integration of bR and the inkjet method holds significant potential for the widespread implementation of highly functional biomaterial-based visual sensors. These sensors have the capability to provide real-time visual information while operating in an energy-efficient manner.

Enhancement method for non-uniform scattering images of turbid underwater based on neural network

Due to the influence of light absorption and scattering in the water medium, underwater images often appear with different degrees of blur, color distortion, and so on. The physical model-based method needs to establish image degradation processes and improve image quality by assuming and estimating complex parameters, which limits the generalization ability and practical application effect of these methods. To solve the above problems, this paper proposes an image enhancement method of non-uniform scattering in turbid underwater based on neural network. First, the high-frequency and low-frequency components of underwater images were obtained through the difference of gaussian (DOG) and bilateral filtering (BF), respectively. Then, for the high-frequency component, pulse coupled neural networks (PCNN) to improve the edge details of the image. For low-frequency components, it is carried out in two steps: an adaptive automatic white balance and polynomial regression (AAWBPR) algorithm is proposed, which can effectively eliminate the color bias under different color temperatures and improve the brightness of images; the adaptive color and contrast enhancement (ACCE) method is used to further enhance the color and contrast of images. Finally, Decom-Net and the improved U-Net method are used to reduce the errors caused by the non-uniform scattering effect, weaken the noise interference, and enhance the overall effect of images. Extensive experiments show that the proposed method can effectively improve the quality of turbid underwater non-uniform scattering images, enhance the contrast of images, reduce the complexity of parameters, and has good generalization performance and practical application effects.

Learnable DoG convolutional filters for microcalcification detection

Difference of Gaussians (DoG) convolutional filters are one of the earliest image processing methods employed for detecting microcalcifications on mammogram images before machine and deep learning methods became widespread. DoG is a blob enhancement filter that consists in subtracting one Gaussian-smoothed version of an image from another less Gaussian-smoothed version of the same image. Smoothing with a Gaussian kernel suppresses high-frequency spatial information, thus DoG can be regarded as a band-pass filter. However, due to their small size and overimposed breast tissue, microcalcifications vary greatly in contrast-to-noise ratio and sharpness. This makes it difficult to find a single DoG configuration that enhances all microcalcifications. In this work, we propose a convolutional network, named DoG-MCNet, where the first layer automatically learns a bank of DoG filters parameterized by their associated standard deviations. We experimentally show that when employed for microcalcification detection, our DoG layer acts as a learnable bank of band-pass preprocessing filters and improves detection performance by 4.86% AUFROC over baseline MCNet and 1.53% AUFROC over state-of-the-art multicontext ensemble of CNNs.

A Novel Day-to-Night Obstacle Detection Method for Excavators Based on Image Enhancement and Multisensor Fusion

Traditional excavator driving relies only on manual observation, resulting in increased hazards in unstructured environments. When the excavator works in a relatively dark environment, there will be potential risks for both the driver and the surrounding pedestrians. In order to address this issue, this study takes the advantages of three different sensors, including infrared cameras, RGB cameras, and lidar sensors, and proposes a novel day-to-night obstacle detection approach by fusing data from multiple sensors. For the dark environment at night, the infrared camera is adopted for the detection task. However, compared with RGB cameras, the infrared camera usually have lower resolutions, making it difficult to be directly applied for the obstacle detection. Therefore, an image enhancement processing method for low-resolution infrared images is developed based on the Difference of Gaussian (DoG). Then, an image recognition method based on YOLO-v5 is proposed to detect images after image enhancement. Finally, a multi-sensor fusion method is suggested to identify the semantic information and 3D coordinates of objects. Experimental studies are carried out to assess image quality and the effectiveness of various object recognition tasks. The results of the experiments demonstrate that our method is capable of not only accurately extracting pedestrian position information from a complicated background environment and realizing timely pedestrian alarms, but also maintaining detection performance in an excavator working environment at night.

An approach for damage detection in noisy environments using DOG multi-scale space and fractal dimension

ABSTRACT The most prominent features of fractal dimension analysis are high sensitivity to damage and instant determination of damage location. However, an intrinsic deficiency is that it is difficult to identify damage in a structure while suppressing noise simultaneously. In addition, the vibration measurement with low-spatial-resolution is also the main factor restricting the accurate damage location. In this paper, a novel damage detection method based on fractal dimension is proposed, in order to provide the high-spatial-resolution vibration information, the phase-based optical flow is introduced, which is a vision-based measurement method that can calculate the structural vibration with high-spatial-resolution. Then, to detect damage in a structure while suppressing noise, an advanced signal analysis method is proposed, with this method, the fractal dimension (FRAC) is mapped into Difference of Gaussian (DOG) multi-scale space, producing the FRAC-DOG to reveal damage-induced singularities in mode shapes. Finally, a data fusion technique is applied to the proposed method to provide a reasonable damage detection result that maximises the elimination of uncertainty. The numerical and experimental results demonstrate that the proposed method can detect damage with high-precision in noisy environments by comparing it with the existing damage detection methods.

Difference-of-Gaussian generative adversarial network for segmenting breast arterial calcifications in mammograms

Breast arterial calcifications (BACs) are amongst the different types of benign calcifications observed on mammograms. BACs have been found to correlate with cardiovascular risk factors, cardiovascular mortality, and coronary artery disease (CAD). Considering that women are recommended to undergo routine screening mammography for the early detection of breast cancer, identifying BACs on mammograms could help identify women at risk of cardiovascular diseases (CVD) without the additional cost or radiation of other tests, such as coronary artery computed tomography (CT). In this paper, we present a difference of Gaussian generative adversarial network (DoG-GAN) model for segmenting BACs in mammograms. It combines the multi-scale difference of Gaussian (DoG) pyramid with the U-net as a generator of the GAN. This approach allows the model to explore image details at each scale and exploit the edge information to effectively segment BACs. We evaluated the performance of our model on a set of synthetic 2D images from digital breast tomosynthesis exams (DBT) collected and prepared for this task. The experimental results show that our model outperforms the state-of-the-art methods.

An Energy-Efficient SIFT Based Feature Extraction Accelerator for High Frame-Rate Video Applications

Visual feature extraction is a key technology of computer vision for intelligent video processing. Efficient feature extraction is a fundamental problem in computer vision applications. Scale-Invariant Feature Transform (SIFT) is one of the most popular feature extraction algorithms because SIFT features are invariant to image scale and rotation and robust to changes in illumination and noise. However, SIFT is a computationally-intensive and power-hungry algorithm, which needs to be accelerated by efficient hardware design to achieve both high-speed feature extraction and high energy efficiency for many high frame-rate video applications at Artificial-intelligent Internet of Things edges. In this work, an energy-efficient SIFT based feature extraction accelerator is proposed. In the Gaussian pyramid and Differences of Gaussian (DoG) pyramid construction process, three design methods are proposed to reduce power consumption and improve information fidelity: a fast and slow dual clock domain design method with a reconfigurable design strategy is proposed to reduce the computation resources; a partial sum reuse design method is proposed to further reduce the computation resources and the amount of computation; a dynamic padding design method is proposed to solve the problem of information loss at image edges and corners after convolution operation. In the keypoint descriptor generation process, an optimized algorithm using circular region and polar coordinates is proposed to parallelize the main orientation assignment and descriptor generation to achieve high-speed processing, while maintaining a comparable matching accuracy with the state-of-the-art designs. The experiment results show that the proposed SIFT hardware accelerator is able to extract features by up to 162 frames per second ( <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$640\times 480$ </tex-math></inline-formula> pixels) under 100 MHz, with the power consumption of 364.26 mW and energy efficiency of 2.25 mJ/frame based on 180 nm technology, which is suitable for many high frame-rate AIoT applications including autonomous driving cars and unmanned aerial vehicles.

DoG-HiT: A novel VLBI multiscale imaging approach

Context. Reconstructing images from very long baseline interferometry (VLBI) data with a sparse sampling of the Fourier domain (uv-coverage) constitutes an ill-posed deconvolution problem. It requires application of robust algorithms, maximizing the information extraction from all of the sampled spatial scales, and minimizing the influence of the unsampled scales on image quality. Aims. We develop a new multiscale wavelet deconvolution algorithm, DoG-HiT, for imaging sparsely sampled interferometric data, which combines the difference of Gaussian (DoG) wavelets and hard image thresholding (HiT). Based on DoG-HiT, we propose a multistep imaging pipeline for analysis of interferometric data. Methods. DoG-HiT applies the compressed sensing approach to imaging by employing a flexible DoG wavelet dictionary, which is designed to adapt smoothly to the uv-coverage. It uses closure properties as data fidelity terms only, initially, and performs nonconvex, nonsmooth optimization by an amplitude-conserving and total-flux-conserving, hard thresholding splitting. DoG-HiT calculates a multiresolution support as a side product. The final reconstruction is refined through self-calibration loops and imaging with amplitude and phase information applied for the multiresolution support only. Results. We demonstrate the stability of DoG-HiT, and benchmark its performance against image reconstructions made with the CLEAN and regularized maximum-likelihood (RML) methods using synthetic data. The comparison shows that DoG-HiT matches the super-resolution achieved by the RML reconstructions and surpasses the sensitivity to extended emission reached by CLEAN. Conclusions. The application of regularized maximum likelihood methods, outfitted with flexible multiscale wavelet dictionaries, to imaging of interferometric data, matches the performance of state-of-the art convex optimization imaging algorithms and requires fewer prior and user-defined constraints.

Detection of Concrete Damage by X-ray CT and Related Non-destructive Elastic Wave Monitoring in Core Test

As a detailed inspection of a concrete structure in service, core samples are usually drilled out and then mechanical properties are measured. In this study, damage estimation of freeze-thawed concrete from concrete-core samples is developed, applying X-ray CT with machine learning method. By the authors, the quantitative damage evaluation of concrete has been proposed by applying acoustic emission and damage mechanics in core concrete. In this study, detection of cracking damage distribution in concrete-core by decision tree method. Prior to the machine learning analysis, distribution of micro-cracks in a concrete-core sample is inspected by helical X-ray computer tomography (CT). Two feature values are set as explanatory variables: luminance value and pixel value after DoG (Difference of Gaussian) filter. The evaluation values converged from the sample size of 12,000 or more. This suggests that one cross section can be estimated with about 50% of the X-ray CT image data by using machine learning. The accuracy rate is 0.723 for the threshold processing and 0.793 for the decision tree method, which is higher for the decision tree method. These results suggest that the proposed method is useful for identifying cracks in concrete by X-ray CT image.

Combination of continuous wavelet transform and genetic algorithm-based Otsu for efficient mass spectrometry peak detection

Mass spectrometry (MS) data is susceptible to random noises and alternating baseline, posing great challenges to spectral peak detection, especially for weak peaks and overlapping peaks. Herein, an efficient peak detection algorithm combining continuous wavelet transform (CWT) and genetic algorithm-based threshold segmentation (denoted as WSTGA) for mass spectrometry was proposed. Firstly, Mexican Hat wavelet was selected as the mother wavelet by comparing the matching degree between the difference of Gaussian (DOG) and different wavelets. Subsequently, the ridges and valleys were identified from 2D wavelet coefficient matrix. Afterward, an improved threshold segmentation method, Otsu method based on genetic algorithm, was introduced to find optimal segmentation threshold and achieve better image segmentation, overcoming the deficiency of traditional Otsu method that cannot handle long-tailed unimodal histograms. Finally, the characteristic peaks were successfully identified by utilizing the ridge-valley lines in wavelet space and original spectrum. Receiver operating characteristic (ROC) curve, area under curve (AUC) and F₁ measure are used as criterions to evaluate performance of peak detection algorithms. Compared with multi-scale peak detection (MSPD) and CWT and image segmentation (CWT-IS) methods, all the results showed that WSTGA can achieve better peak detection. More importantly, the experimental results from MALDI-TOF spectra demonstrated that WSTGA can effectively detect more weak peaks and overlapping peaks while maintaining a lower false peak detection rate than MSPD and CWT-IS methods, indicating its great advantages in characteristic peak identification.

Robust data hiding scheme through distinct keypoint selection exploiting modified Bilateral-Laplacian SIFT with encoding pipeline

SIFT (Scale Invariant Feature Transform) has recently been used to secure multimedia materials from geometric and non-geometric attacks. SIFT is commonly used in Computer Vision to identify and describe local features in images using the Difference of Gaussian (DoG) method, which has a high processing cost and a strong affinity towards the edge, resulting in a low contrast point. By introducing modified SIFT (mSIFT), which selects the most distinguishable key point with their feature descriptors, an attempt has been made to construct a rotation and scaling resistant data hiding strategy in this study. Unlike SIFT, a Bilateral-Laplacian filter was employed to determine the extrema, which took into account both geometric and photometric distance to reduce noise and retain the edge. This adds to the benefits of the proposed scheme’s resistance against geometric attacks, even when the tampering rate is large. Furthermore, during data embedding, the suggested technique achieves excellent security by utilizing Arnold chaotic function, pseudorandom sequence generator, and a newly created unique encoding pipeline. The proposed technique is secure and resistant to various steganographic attacks while preserving good visual quality, according to experimental results and analysis, as well as comparisons with existing state-of-the-art methods.

Dexterous holographic trapping of dark-seeking particles with Zernike holograms.

The intensity distribution of a holographically-projected optical trap can be tailored to the physical properties of the particles it is intended to trap. Dynamic optimization is especially desirable for manipulating dark-seeking particles that are repelled by conventional optical tweezers, and even more so when dark-seeking particles coexist in the same system as light-seeking particles. We address the need for dexterous manipulation of dark-seeking particles by introducing a class of "dark" traps created from the superposition of two out-of-phase Gaussian modes with different waist diameters. Interference in the difference-of-Gaussians (DoG) trap creates a dark central core that is completely surrounded by light and therefore can trap dark-seeking particles rigidly in three dimensions. DoG traps can be combined with conventional optical tweezers and other types of traps for use in heterogeneous samples. The ideal hologram for a DoG trap being purely real-valued, we introduce a general method based on the Zernike phase-contrast principle to project real-valued holograms with the phase-only diffractive optical elements used in standard holographic optical trapping systems. We demonstrate the capabilities of DoG traps (and Zernike holograms) through experimental studies on high-index, low-index and absorbing colloidal particles dispersed in fluid media.

Ultra-fast detail enhancement for a short-wave infrared image.

Image detail enhancement is critical to the performance of short-wave infrared (SWIR) imaging systems, especially to the long-range systems. However, the existing high-performance infrared (IR) image enhancement methods typically have difficulty in meeting the requirements of the imaging system with high resolution and high frame rate. In this paper, we propose an ultra-fast and simple SWIR image detail enhancement method based on the difference of Gaussian (DoG) filter and plateau equalization. Our method consists of efficient edge detail information extraction and histogram equalization. The experimental results demonstrated that the proposed method achieves outstanding enhancement performance with a frame rate around 50 fps for 1280×1024 SWIR images.

Scale-Invariant Localization of Electric Vehicle Charging Port via Semi-Global Matching of Binocular Images

Automatic charging for electric vehicles has broad development prospects for meeting the personalized service experience of users while overcoming the inherent safety hazards. An identification and positioning approach suitable for engineering applications is the key to promoting automatic charging. In this paper, a low-cost, high-precision method to identify and position charging ports based on SIFT and SGBM is proposed. The feature extraction approach based on SIFT is adopted to produce the difference of Gaussian (DOG) for scale space construction, and the feature matching algorithm with nearest-neighbor search, which is a kind of machine learning, is utilized to yield the map set of matching points. In addition, the disparity calculation is conducted with a semi-global matching algorithm to obtain high-precision positioning results for the charging port. In order to verify the feasibility of the method, a complete identification and positioning experiment of charging port was carried out based on OpenCV and MATLAB.