Mean IoU Research Articles

Deep labeller: automatic bounding box generation for synthetic violence detection datasets

Manually labelling datasets for training violence detection systems is time-consuming, expensive, and labor-intensive. Mind wandering, boredom, and short attention span can also cause labelling errors. Moreover, collecting and distributing sensitive images containing violence has ethical implications. Automation is the future for labelling sensitive image datasets. Deep labeller is a two-stage Deep Learning (DL) method that uses pre-trained DL object detection methods on MS-COCO for automatic labelling. The Deep Labeller method labels violent and nonviolent images in WVD and USI. In stage 1, WVD generates weak labels using synthetic images. In stage 2, the Deep labeller method is retrained on weak labels. USI dataset is used to test our method on real-world violence. Deep labeller generated weak and strong labels with an IoU of 0.80036 in stage 1 and 0.95 in stage 2 on the WVD. Automatically generated labels. To test our method’s generalisation power, violent and nonviolent image labels on USI dataset had a mean IoU of 0.7450.

PDDS: A software for the early diagnosis of Parkinson’s disease from MRI and DaT scan images using detection and segmentation algorithms

Parkinson’s disease is a common neurodegenerative disorder that affects millions of people worldwide. Medical imaging techniques such as MRI and DaTScan can help diagnose PD by showing the changes in the brain structures related to the disease, such as the substantia nigra and the striatum. In this paper, we propose a software called PDDS (Parkinson’s Disease Diagnosis Software) that can automatically detect and segment the deep brain regions from MRI and DaTScan images using advanced deep learning models. Our software uses YOLO and an ensemble of UNETs to accurately locate the regions of interest (ROIs) from T2-weighted MRI and DaTScan images respectively and use them for quantitative assessment and decision support. The ensemble approach uses a weighted average ensemble which achieves the highest mean IOU after a grid search technique was implemented to get the appropriate weights. We used mosaic data augmentation to improve the generalization ability of our models by exposing them to objects in different contexts. The proposed architecture was trained and validated using the PPMI and IXI datasets and the detection results show that the YOLOv7x model achieves a mAP_0.5:0.95 of 70.39 % for DaTScan images and a mAP_0.5:0.95 of 64.16 % for MRI images, outperforming previous methods. For the segmentation results, our ensemble UNet model with different backbones achieves a mean IOU of 70.02 % for DaTScan images and 54.31 % for MRI images. According to our findings, our approach is a competitive automatic detection and segmentation method and can be applied to clinically challenging medical imaging problems.

One-stage 3D profile-based pavement crack detection and quantification

To improve the robustness and efficiency of pavement crack inspection tasks, a one-stage pavement crack detection and quantification method is proposed to directly extract crack geometry from 3D pavement profile. A low-cost stereo imaging system is implemented for high-resolution 3D reconstruction. Based on the generated 3D profile, an algorithm that integrates crack detection and quantification is developed to obtain the crack map, length, width, and depth simultaneously. The results show that the proposed method is robust to surrounding noises and has no demand for prior data. Compared with deep learning-based crack detection methods, the mean IoU of the proposed method is 6.23% higher than that of state-of-the-art methods. The proposed method outperforms existing methods in terms of crack length and width measurement. Moreover, the proposed method automatically measures crack depth with a correlation coefficient R2 of 0.9679 against the manual measurement.

Semi-supervised nuclei segmentation based on multi-edge features fusion attention network.

The morphology of the nuclei represents most of the clinical pathological information, and nuclei segmentation is a vital step in current automated histopathological image analysis. Supervised machine learning-based segmentation models have already achieved outstanding performance with sufficiently precise human annotations. Nevertheless, outlining such labels on numerous nuclei is extremely professional needing and time consuming. Automatic nuclei segmentation with minimal manual interventions is highly needed to promote the effectiveness of clinical pathological researches. Semi-supervised learning greatly reduces the dependence on labeled samples while ensuring sufficient accuracy. In this paper, we propose a Multi-Edge Feature Fusion Attention Network (MEFFA-Net) with three feature inputs including image, pseudo-mask and edge, which enhances its learning ability by considering multiple features. Only a few labeled nuclei boundaries are used to train annotations on the remaining mostly unlabeled data. The MEFFA-Net creates more precise boundary masks for nucleus segmentation based on pseudo-masks, which greatly reduces the dependence on manual labeling. The MEFFA-Block focuses on the nuclei outline and selects features conducive to segment, making full use of the multiple features in segmentation. Experimental results on public multi-organ databases including MoNuSeg, CPM-17 and CoNSeP show that the proposed model has the mean IoU segmentation evaluations of 0.706, 0.751, and 0.722, respectively. The model also achieves better results than some cutting-edge methods while the labeling work is reduced to 1/8 of common supervised strategies. Our method provides a more efficient and accurate basis for nuclei segmentations and further quantifications in pathological researches.

OC-035 AUTOMATION OF HERNIA VOLUMES - SAC DETECTION AND HERNIA VOLUMETRY USING COMPUTER VISION AND DEEP LEARNING - PUSHING FRONTIERS

Abstract Aim To identify the sac and defect using deep learning algorithms and to compute hernial sac volume in an automated fashion. Materials and methods 70 ventral hernia CT scans were used. Slices with hernia defect and sac were annotated using computer vision tools. 80% of the data was used for training and validation. 20% data was used for testing. We used a combination of 3D reconstruction algorithms and deep learning to train the models. The model output was compared to manual annotation by experts. Results The testing data showed that our novel algorithm was able tp estimate the defect with an accuracy of 90%. The mean IOU for detection of hernia sac was 88% and volume estimation accuracy was 85%. Conclusion A combination of supervised deep learning and 3D volumetry will increase the accuracy of volumetric indices and shorten the volume estimation time. In most parts of the world, the hernia sac volume is not calculated by the radiologist and the surgeon is left to estimate it using rudimentary means such as visual estimation. Automated hernia volumetry will play a major role in standardising hernia reports. We believe this is a pivotal step in creating actionable radiology reports that will help the surgeon assess the complexity and plan surgery.

Deep ensemble learning for quantitative geological fracture analysis using borehole televiewer images

In the geological survey of hydraulic engineering projects, the Borehole televiewer (BHTV) technique is applied widely. The BHTV images are usually interpreted by engineers. To improve the automatic level of geological images analysis and decrease the subjective influence of the engineers, a fracture quantitative approach based on deep ensemble learning is proposed. In the research, the BHTV images' segmentation using the ensemble model is the basis of the fracture quantitative analysis. Therefore, the accuracy of segmentation can not be ignored. In the experimental study, the different deep learning architectures, UNet, SegNet, and FRRN based on VGG and ResNet, are employed. The evaluation result shows that the IoU values of the single models are in the range of 83.96%–84.36%, while the ensemble model outperforms all of them with a mean IoU of 86.36% and ranks top-3 for the all testing images. Through comparison of the segmentation for the single image, it is evident that the ensemble model is effective for noises removal, which improves the accuracy and efficiency of fracture parameter quantitative analysis. Finally, the thinning method and sine function based on Hough Transform (HT) and the generalized least squares (GLS) are adopted to fit the fracture curve. The calculation result of the dip azimuth, dip angle, and mean thickness also validates the effectiveness of the ensemble model.

Automatic Analysis of Transverse Musculoskeletal Ultrasound Images Based on the Multi-Task Learning Model

Musculoskeletal ultrasound imaging is an important basis for the early screening and accurate treatment of muscle disorders. It allows the observation of muscle status to screen for underlying neuromuscular diseases including myasthenia gravis, myotonic dystrophy, and ankylosing muscular dystrophy. Due to the complexity of skeletal muscle ultrasound image noise, it is a tedious and time-consuming process to analyze. Therefore, we proposed a multi-task learning-based approach to automatically segment and initially diagnose transverse musculoskeletal ultrasound images. The method implements muscle cross-sectional area (CSA) segmentation and abnormal muscle classification by constructing a multi-task model based on multi-scale fusion and attention mechanisms (MMA-Net). The model exploits the correlation between tasks by sharing a part of the shallow network and adding connections to exchange information in the deep network. The multi-scale feature fusion module and attention mechanism were added to MMA-Net to increase the receptive field and enhance the feature extraction ability. Experiments were conducted using a total of 1827 medial gastrocnemius ultrasound images from multiple subjects. Ten percent of the samples were randomly selected for testing, 10% as the validation set, and the remaining 80% as the training set. The results show that the proposed network structure and the added modules are effective. Compared with advanced single-task models and existing analysis methods, our method has a better performance at classification and segmentation. The mean Dice coefficients and IoU of muscle cross-sectional area segmentation were 96.74% and 94.10%, respectively. The accuracy and recall of abnormal muscle classification were 95.60% and 94.96%. The proposed method achieves convenient and accurate analysis of transverse musculoskeletal ultrasound images, which can assist physicians in the diagnosis and treatment of muscle diseases from multiple perspectives.

A Foreground Prototype-Based One-Shot Segmentation of Brain Tumors.

The potential for enhancing brain tumor segmentation with few-shot learning is enormous. While several deep learning networks (DNNs) show promising segmentation results, they all take a substantial amount of training data in order to yield appropriate results. Moreover, a prominent problem for most of these models is to perform well in unseen classes. To overcome these challenges, we propose a one-shot learning model to segment brain tumors on brain magnetic resonance images (MRI) based on a single prototype similarity score. With the use of recently developed few-shot learning techniques, where training and testing are carried out utilizing support and query sets of images, we attempt to acquire a definitive tumor region by focusing on slices containing foreground classes. It is unlike other recent DNNs that employed the entire set of images. The training of this model is carried out in an iterative manner where in each iteration, random slices containing foreground classes of randomly sampled data are selected as the query set, along with a different random slice from the same sample as the support set. In order to differentiate query images from class prototypes, we used a metric learning-based approach based on non-parametric thresholds. We employed the multimodal Brain Tumor Image Segmentation (BraTS) 2021 dataset with 60 training images and 350 testing images. The effectiveness of the model is evaluated using the mean dice score and mean IoU score. The experimental results provided a dice score of 83.42 which was greater than other works in the literature. Additionally, the proposed one-shot segmentation model outperforms the conventional methods in terms of computational time, memory usage, and the number of data.

Brain tumor segmentation by cascaded multiscale multitask learning framework based on feature aggregation

Brain tumor analysis in MRI images is a significant and challenging issue because misdiagnosis can lead to death. Diagnosis and evaluation of brain tumors in the early stages increase the probability of successful treatment. However, the complexity and variety of tumors, shapes, and locations make their segmentation and classification complex. Numerous researchers have proposed brain tumor segmentation and classification methods in this regard. This paper presents an approach that simultaneously segments and classifies brain tumors in MRI images using a framework that contains MRI image enhancement and tumor region detection. Eventually, a network based on a multitask learning approach is proposed. The proposed network, called Multiscale Cascaded Multitask Network, is based on a multitask learning approach containing segmentation and classification tasks. A multiscale approach and cascade approach in layers of encoder and decoder have been applied to improve segmentation accuracy in the proposed network. In addition, to increase the classification accuracy, a feature aggregation module has been introduced that integrates different levels of features to better tumor type classification. Simultaneously learning the two tasks of segmentation and classification, along with applying the mentioned approaches, has improved the results in both tasks. Subjective and objective results indicate that the segmentation and classification results based on evaluation metrics are better or comparable to the state-of-the-art. Our proposed method has reached 96.27 and 95.88 for DCS and mean IoU, respectively, for segmentation and 97.988 accuracies for classification.

RFS-Net: Railway Track Fastener Segmentation Network With Shape Guidance

The fastener is one of the main components of a rail track system. In recent years, deep learning methods such as image segmentation have greatly boosted the fastener state detection process. However, there is still a need to improve the segmentation accuracy and speed, especially for the fasteners in complex environments. To handle this problem, a fast and accurate fastener semantic segmentation network named RFS-Net is proposed based on shape guidance, which can offer a better speed/accuracy trade-off performance via a very shallow architecture. Specifically, in the encoder, a two-stream structure (i.e., regular stream and shape stream) that processes the fastener and shape image in parallel is introduced. The shape image is created based on the geometric structure of the fastener, and it is served as input to the shape stream to guide the segmentation of the fastener. The decoder integrates deep features from the two-stream encoder and then recovers the shape information by the shape attention blocks with skipping connections. We provide two versions of RFS-Net: RFS-Net_S (1.0M, 1014FPS) and RFS-Net_L (12.01M, 453FPS) on the NVIDIA RTX 3060. Experimental results demonstrate the effectiveness of our method by achieving a promising trade-off between accuracy and inference speed. In particular, our method is faster and more accurate on a challenging dataset, from fast modes: 1014 FPS for RFS-Net_S versus 724 FPS for Segmenter, to high-quality segmentation: better performance than STDC with nearly one percent (92.36% versus 91.48% Mean IoU score).

TransFusionNet: Semantic and Spatial Features Fusion Framework for Liver Tumor and Vessel Segmentation Under JetsonTX2.

Liver cancer is one of the most common malignant diseases worldwide. Segmentation and reconstruction of liver tumors and vessels in CT images can provide convenience for physicians in preoperative planning and surgical intervention. In this paper, we introduced a TransFusionNet framework, which consists of a semantic feature extraction module, a local spatial feature extraction module, an edge feature extraction module, and a multi-scale feature fusion module to achieve fine-grained segmentation of liver tumors and vessels. In addition, we applied the transfer learning approach to pre-train using public datasets and then fine-tune the model to further improve the fitting effect. Furthermore, we proposed an intelligent quantization scheme to compress the model weights and achieved high performance inference on JetsonTX2. The TransFusionNet framework achieved mean IoU of 0.854 in vessel segmentation task, and achieved mean IoU of 0.927 in liver tumor segmentation task. When profiling the Computational Performance of the quantized inference, our quantized model achieved 4TFLOPs on Node with NVIDIA RTX3090 and 132GFLOPs on JetsonTX2. This unprecedented segmentation effect solves the accuracy and performance bottleneck of automated segmentation to a certain extent.

Improving GAN Learning Dynamics for Thyroid Nodule Segmentation.

Thyroid nodules are lesions requiring diagnosis and follow-up. Tools for detecting and segmenting nodules can help physicians with this diagnosis. Besides immediate diagnosis, automated tools can also enable tracking of the probability of malignancy over time. This paper demonstrates a new algorithm for segmenting thyroid nodules in ultrasound images. The algorithm combines traditional supervised semantic segmentation with unsupervised learning using GANs. The hybrid approach has the potential to upgrade the semantic segmentation model's performance, but GANs have the well-known problems of unstable learning and mode collapse. To stabilize the training of the GAN model, we introduce the concept of closed-loop control of the gain on the loss output of the discriminator. We find gain control leads to smoother generator training and avoids the mode collapse that typically occurs when the discriminator learns too quickly relative to the generator. We also find that the combination of the supervised and unsupervised learning styles encourages both low-level accuracy and high-level consistency. As a test of the concept of controlled hybrid supervised and unsupervised semantic segmentation, we introduce a new model named the StableSeg GAN. The model uses DeeplabV3+ as the generator, Resnet18 as the discriminator, and uses PID control to stabilize the GAN learning process. The performance of the new model in terms of IoU is better than DeeplabV3+, with mean IoU of 81.26% on a challenging test set. The results of our thyroid nodule segmentation experiments show that StableSeg GANs have flexibility to segment nodules more accurately than either comparable supervised segmentation models or uncontrolled GANs.

Ultrasound detection of subquadricipital recess distension

Joint bleeding is a common occurrence for people with hemophilia and, if untreated, can result in hemophilic arthropathy. Ultrasound imaging has recently emerged as an effective tool to diagnose the distension of a joint recess caused by joint bleeding. However, no computer-aided diagnosis tool exists to support the practitioner in the diagnosis process. This paper addresses the problem of automatically detecting the subquadricipital recess of the knee and assessing whether it is distended in ultrasound images collected from patients with hemophilia. After framing the problem, we propose two different approaches: the first one adapts a one-stage object detection algorithm for the task, while the second one is a multi-task approach with a classification and a detection branch. The experimental evaluation, conducted with 483 annotated images, shows that the solution based on object detection alone has a balanced accuracy score of 0.74 with a mean IoU value of 0.66, while the multi-task approach has a higher balanced accuracy value (0.78) at the cost of a slightly lower mean IoU value.

A shape-attention Pivot-Net for identifying central pivot irrigation systems from satellite images using a cloud computing platform: an application in the contiguous US

ABSTRACT Forty percent of global food production relies upon irrigation, which accounts for 70% of total global freshwater use. Thus, the mapping of cropland irrigation plays a significant role in agricultural water management and estimating food production. However, current spaceborne irrigated cropland mapping is highly reliant upon its spectral behavior, which often has high uncertainty and lacks information about the method of irrigation. Deep learning (DL) allows for the classification of irrigated cropland according to unique spatial patterns, such as the central pivot irrigation system (CPIS). But convolutional neural networks (CNNs) are usually biased toward color and texture features, a spatial transferable and accurate CPIS identification model is lacking owing to previous model seldom involves the round shapes of CPIS, which is usually key to distinguishing CPIS. To address this lack, we proposed a shape attention neural network by integrating spatial-attention gate, residual block, and multi-task learning, Pivot-Net, to incorporate shape information identify CPIS in satellite imagery. Specifically, we employed CPIS in Kansas to train our model using Sentinel-2 and Landsat-8 optical images. We found that Pivot-Net is superior to seven state-of-the-art semantic segmentation models on second-stage validation. We also evaluated the performance of Pivot-Net at three validation sites, which had an average F1-score and mean IOU of 90.68% and 90.45%, respectively, which further demonstrated the high accuracy of the proposed model. Moreover, to show that the proposed Pivot-Net can map CPIS at the country scale, we generated the first CPIS map at 30 m for the contiguous US using a cloud computing platform and our Pivot-Net model. The total CPIS area for the contiguous US was 61,094 km2 in 2018, which comprised 26.22% of all irrigated lands. Our results can be accessed at https://tianfyou.users.earthengine.app/view/cpisus. Therefore, the proposed shape-attention Pivot-Net demonstrates the ability to classify CPIS at large spatial scales and are feasible to map CPIS at national scales.

AInBody: Are you in shape? - An integrated deep learning model that tracks your body measurement

This paper presents AInBody, a novel deep learning-based body shape measurement solution. We have devised a user-centered design that automatically tracks the progress of the body by adequately integrating various methods, including human parsing, instance segmentation, and image matting. Our system guides a user's pose when taking photos by displaying the outline of the latest picture of the user, divides the human body into several parts, and compares before and after photos of the body part level. The parsing performance has been improved through an ensemble approach and a denoising phase in our main module, Advanced Human Parser. In evaluation, the proposed method is 0.1% to 4.8% better than the other best-performing model in average precision in 3 out of 5 parts, and 1.4% and 2.4% superior in mAP and mean IoU, respectively. Furthermore, the inference time of our framework takes approximately three seconds to process one HD image, demonstrating that our structure can be applied to real-time applications.

Scattered Mountainous Area Building Extraction From an Open Satellite Imagery Dataset

Building extraction from satellite images has been a hot research topic in the field of remote sensing image analysis. Most of the related studies are focusing on the urban areas with dense populations, while solutions for underpopulated mountainous areas are still missing. Given the scarcity of the research materials, it is still an important topic for applications like mountain hazard damage management. To fill this gap, we present a new dataset for scattered mountainous area building segmentation, consisting of the manual labels, the coordinates of latitude and longitude for 2125 satellite images of 303 diverse human settlements in the mountainous areas of southwest China. Compared with the public remote sensing image datasets, our dataset is very challenging with relatively low resolution (2.2m/pixel) and small-object, blurry-boundary and high-imbalance features. In this paper, we propose a novel Copy-Fusion data augmentation strategy, and designed a VGG-16+U-Net with Dice Focal loss to address the difficulties in this building extraction task. Experiment results have shown that our approach has achieved 81.79% mean IoU and outperformed the baseline building segmentation methods like HRNet, DeeplabV3+, and U-net+ResNet by 6.29%. Our dataset and model will be available at https://github.com/AngCV/SMAB_DATASET.

DDFormer: A Dual-Domain Transformer for Building Damage Detection Using High-Resolution SAR Imagery

Earthquakes are catastrophic in terms of damage to buildings. Synthetic aperture radar (SAR) has emerged as an effective tool to respond to seismic hazards. However, pre- and post-event high-resolution data are not always available for the affected areas, and the complex geometric properties of buildings pose a challenge to building damage detection. Therefore, this letter proposes the Dual-Domain Transformer (DDFormer) semantic segmentation model for damaged buildings detection using a single post-earthquake high-resolution SAR image. The difference between intact and collapsed building features is enhanced by adaptive frequency and spatial modules. Taking the 2023 Turkey earthquake as an example, the experiments are conducted on two high-resolution co-polarized SAR data (Capella and GF-3). The DDFormer achieves optimal detection accuracy with mean IOU (mIOU) and F-Score of 81.81% and 90%, respectively. In addition, our results are in high consistent with the Turkey Earthquake Report published by Microsoft with a correlation coefficient of 0.626. The above experiments demonstrate the robustness and effectiveness of DDFormer.

PST: Plant segmentation transformer for 3D point clouds of rapeseed plants at the podding stage

Segmentation of plant point clouds to obtain high-precise morphological traits is essential for plant phenotyping. Although the fast development of deep learning has boosted much research on segmentation of plant point clouds, previous studies mainly focus on the hard voxelization-based or down-sampling-based methods, which are limited to segmenting simple plant organs. Segmentation of complex plant point clouds with a high spatial resolution still remains challenging. In this study, we proposed a deep learning network plant segmentation transformer (PST) to achieve the semantic segmentation of rapeseed plants point clouds acquired by handheld laser scanning (HLS) with the high spatial resolution, which can characterize the tiny siliques as the main traits targeted. PST is composed of: (i) a dynamic voxel feature encoder (DVFE) to aggregate the point features with the raw spatial resolution; (ii) the dual window sets attention blocks to capture the contextual information; and (iii) a dense feature propagation module to obtain the final dense point feature map. We then integrated PST with an instance segmentation head in the point grouping network (PointGroup) and developed PST-PointGroup (PG) to achieve the instance segmentation of the siliques. The results proved that PST and PST-PG achieved superior performance in semantic and instance segmentation tasks. For the semantic segmentation, the mean IoU, mean Precision, mean Recall, mean F1-score, and overall accuracy of PST were 93.96%, 97.29%, 96.52%, 96.88%, and 97.07%, achieving an improvement of 7.62, 3.28, 4.8, 4.25, and 3.88 percentage points compared to the second-best state-of-the-art network position adaptive convolution (PAConv). For instance segmentation, PST-PG reached 89.51%, 89.85%, 88.83% and 82.53% in mCov, mWCov, mPerc90, and mRec90, achieving an improvement of 2.93, 2.21, 1.99, and 5.9 percentage points compared to the original instance segmentation network PointGroup. This study extends the phenotyping of rapeseed plants in an end-to-end way and proves that the deep-learning-based point cloud segmentation method has a great potential for resolving dense plant point clouds with complex morphological traits.

Link-RGBD: Cross-Guided Feature Fusion Network for RGBD Semantic Segmentation

Significant advances have been made in the development of deep convolutional networks for RGBD semantic segmentation. However, the full extraction and utilization of depth information to assist semantic segmentation remain challenging. In this study, we propose the Link-RGBD module to fuse RGB and depth features using an innovative interactive attention mechanism to augment the representations of the objects of interest. The attention map gotten from RGB and depth branch is regarded as a weight for value (v) in the other branch in order to achieve cross-fusion. This enhances the coordinates and feature information and augments the representations of the objects of interest. We evaluated the effectiveness of the proposed method and achieved competitive results on the challenging SUN RGB-D and NYUDv2 semantic segmentation benchmarks. Its pixel accuracy is 0.2% higher than that of DeepLabv3+, reaching 83.1% on the SUN RGB-D benchmark. Moreover, its mean IoU score is 0.3% higher than that of DeepLabv3+, reaching 49.5% on the NYUDv2 benchmark. We further constructed a robot test system to evaluate the performance of the proposed method in a real scenario. The results show that the proposed method can support the completion of the moving forward and grasping task.

Relevancy between Objects Based on Common Sense for Semantic Segmentation

Research on image classification sparked the latest deep-learning boom. Many downstream tasks, including semantic segmentation, benefit from it. The state-of-the-art semantic segmentation models are all based on deep learning, and they sometimes make some semantic mistakes. In a semantic segmentation dataset with a small number of categories, images are often collected from a single scene, and there is a close semantic connection between any two categories. However, in the semantic segmentation dataset collected from multiple scenes, two categories may be irrelevant. The probability of objects in one category appearing next to objects in other categories is different, which is the basis of the paper. Semantic segmentation methods need to solve two problems of positioning and classification. This paper is dedicated to correcting those clearly wrong classifications that are contrary to reality. Specifically, we first calculate the relevancy between different class pairs. Then, based on this knowledge, we infer the category of a connected component according to the relationships of this connected component with its surrounding connected components and correct the obviously wrong classifications made by a deep learning semantic segmentation model. Several well-performing deep learning models are experimented on two challenging public datasets in the field of semantic image segmentation. Our proposed method improves the performance of UPerNet, OCRNet and SETR from 40.7%, 43% and 48.64% to 42.07%, 44.09% and 49.09% mean IoU on the ADE20K validation set, and the performance of PSPNet, DeepLabV3 and OCRNet from 37.26%, 37.3% and 39.5% to 38.93%, 38.95% and 40.63% mean IoU on the COCO-Stuff dataset, which shows the effectiveness of the method.