Semantic Image Segmentation with Deep Convolutional Neural Networks and Quick Shift

Zhang, Sanxing; Ma, Zhenhuan; Zhang, Gang; Leí, Tao; Zhang, Rui; Cui, Yi

doi:10.3390/sym12030427

Cited by 38 publications

(24 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Deeplabv3+ with Xception backbone showed the best performance over the PASCAL VOC 2012 dataset, but the accuracy of FPN was higher. In the former, atrous convolutional layers are particularly suitable for capturing contextual information, but much of the edge information is lost, reducing its ability to detect small objects (Guo et al 2019;S. Zhang et al 2020).…”

Section: Discussionmentioning

confidence: 99%

“…Two studies used CNN models to estimate percent severity of foliar symptoms using images obtained under controlled environment (Lin et al 2019) or field experiments (Tusubira et al 2020), both making use of semantic segmentation approaches. Also known as pixel-level classification, semantic segmentation models have been tested in a wide range of classification tasks using images of medical (Borne et al 2020;Y. Zhang et al 2020) or remote sensing (Guo et al 2019;Krestenitis et al 2019) applications, with superior performance compared with the traditional deep learning methods.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Deep learning models for semantic segmentation and automatic estimation of severity of foliar symptoms caused by diseases or pests

Gonçalves¹,

Pinto²,

Queiroz³

et al. 2020

Preprint

View full text Add to dashboard Cite

Measures of percent severity of visible symptoms or injuries caused by diseases or insect pests on plant organs are essential in plant health research. Current color thresholding digital imaging-methods are generally more accurate and reliable than visual estimates. However, these methods perform poorly when scene illumination and background are not uniform, conditions that can be overcome by convolutional neural networks (CNN) for semantic segmentation. In this study, we trained five CNN models for pixel level predictions in images of individual leaves exhibiting necrotic lesions and/or yellowing caused by the insect pest coffee leaf miner (CLM), and two fungal diseases: soybean rust (SBR) and wheat tan spot (WTS). Training was performed in 80% of images annotated for three classes: leaf background (B), healthy leaf (H) and injured leaf (I). Precision, recall, and Intersection over Union (IoU) metrics in the test image set were highest for B, followed by H and I classes, irrespective of the model. When the pixel-level predictions were used to estimate percent severity, Feature Pyramid Network (FPN), Unet and DeepLabv3+ (Xception) performed the best: concordance coefficients were greater than 0.95, 0.96 and 0.98 for CLM, SBR and WTS datasets, respectively. The other three models tended to misclassify healthy pixels as injured, leading to overestimation of percent severity. The accuracy of the predictions by CNN models were comparable with those obtained using a standard commercial software which requires manual adjustments that slows the process.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Deep learning models for semantic segmentation and automatic estimation of severity of foliar symptoms caused by diseases or pests

Gonçalves¹,

Pinto²,

Queiroz³

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Simple Linear Iterative Clustering (SLIC) is a superpixel segmentation algorithm proposed by Achanta et al SLIC is a k-Means based algorithm that clusters neighboring pixels considering color and coordinate information [23]. When SLIC is used as a superpixel preprocessing step, it provides computational speed, memory savings [24].…”

Section: Simple Linear Iterative Clustering (Slic) Methodsmentioning

confidence: 99%

Automatic Cell Nucleus Segmentation Using Superpixels and Clustering Methods in Histopathological Images

Mendi

Budak

2021

Balkan Journal of Electrical and Computer Engineering

View full text Add to dashboard Cite

It is seen that there is an increase in cancer and cancer-related deaths day by day. Early diagnosis is vital for the early treatment of the cancerous area. Computer-aided programs allow for the early diagnosis of unhealthy cells that specialist pathologists diagnose due to efforts.In this study, clustering and superpixel segmentation techniques were used to detect cell nuclei in high-resolution histopathology images automatically. As a result of the study, the successful performances of the segmentation algorithms were analyzed and evaluated. It is seen that better success is obtained in the Watershed and FCM algorithms in highresolution histopathological images used. Quickshift and SLIC methods gave better results in terms of precision. It is seen that there are k-Means and FCM algorithms that provide the best performance in F measure (F-M), and the correct negative rate (TNR) is more successful in Quickshift, k-Means, and SLIC methods.

show abstract

“…The decoder learns the information lost in the downsampling and improves the segmentation effect. The DeepLabV1 proposed in [25] uses hole convolution to reduce the size of the feature map while keeping the resolution unchanged. And at the end of the network, the conditional random field (CRF) model is used to restore the edge information and achieve accurate positioning.…”

Section: Overall Structure and Basic Conceptsmentioning

confidence: 99%

Enet Semantic Segmentation Combined with Attention Mechanism

Bai

2021

Preprint

View full text Add to dashboard Cite

Image semantic segmentation is one of the core tasks of computer vision. It is widely used in fields such as unmanned driving, medical image processing, geographic information systems and intelligent robots. Aiming at the problem that the existing semantic segmentation algorithm ignores the different channel and location features of the feature map and the simple method when the feature map is fused, this paper designs a semantic segmentation algorithm that combines the attention mechanism. Firstly, dilated convolution is used, and a smaller downsampling factor is used to maintain the resolution of the image and obtain the detailed information of the image. Secondly, the attention mechanism module is introduced to assign weights to different parts of the feature map, which reduces the accuracy loss. The design feature fusion module assigns weights to the feature maps of different receptive fields obtained by the two paths, and merges them together to obtain the final segmentation result. Finally, through experiments, it was verified on the Camvid, Cityscapes and PASCAL VOC2012 datasets. Mean intersection over union (MIoU) and mean pixel accuracy (MPA) are used as metrics. The method in this paper can make up for the loss of accuracy caused by downsampling while ensuring the receptive field and improving the resolution, which can better guide the model learning. And the proposed feature fusion module can better integrate the features of different receptive fields. Therefore, the proposed method can significantly improve the segmentation performance compared to the traditional method.

show abstract

Semantic Image Segmentation with Deep Convolutional Neural Networks and Quick Shift

Cited by 38 publications

References 19 publications

Deep learning models for semantic segmentation and automatic estimation of severity of foliar symptoms caused by diseases or pests

Deep learning models for semantic segmentation and automatic estimation of severity of foliar symptoms caused by diseases or pests

Automatic Cell Nucleus Segmentation Using Superpixels and Clustering Methods in Histopathological Images

Enet Semantic Segmentation Combined with Attention Mechanism

Contact Info

Product

Resources

About