Automatic Road Crack Detection Using Random Structured Forests

Shi, Yong; Cui, Limeng; Qi, Zhiquan; Fan, Meng; Chen, Zhen‐Song

doi:10.1109/tits.2016.2552248

Cited by 985 publications

(656 citation statements)

References 39 publications

Supporting

Mentioning

654

Contrasting

Unclassified

Order By: Relevance

“…We use a direct comparison between the PGT and the segmented result; for comparison, a 2 pixels margin was used in [2] and a larger margin in [3]. The performance evaluation involves the following four categories of pixels: true positive pixels (TP), false positive pixels (FP), false negative pixels (FN) and true negative pixels (TN).…”

Section: Assessment Protocolmentioning

confidence: 99%

“…In Fig. 2, the five algorithms are performed on the pavement image 462×690 pixels in size, which has been used in [2][3][4] for testing automatic crack detection algorithms. The black line represents the crack (with the thickness accounted for) and the query area in grey displays the visited pixels by the algorithms The results shown in Fig.…”

Section: A Test On a Pavement Image Samplementioning

confidence: 99%

“…A lot of image processing techniques have been developed to computerize the survey of road surface conditions and for instance, to automatically detect the cracks within 2D pavement images, e.g., [1][2][3][4][5][6] and the references therein. However, none of the existing algorithms actually affords the ideal segmentation which may serve as the references for others.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Pseudo-ground truth data collection on pavement images

Baltazart¹,

Yang²,

Nicolle³

et al. 2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

Abstract-The performance assessment of automatic crack detection algorithms within pavement images requires beforehand to establish a reference image, namely, the pseudoground truth image (PGT). In this context, this paper presents some existing pseudo-ground truth (PGT) data collection techniques which rely on image processing techniques. The processing of five Single Pair Shortest Path (SPSP) algorithms which are devoted to this aim are illustrated in terms of running time and segmentation accuracy on a pavement image.

show abstract

Section: Assessment Protocolmentioning

confidence: 99%

Section: A Test On a Pavement Image Samplementioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Pseudo-ground truth data collection on pavement images

Baltazart¹,

Yang²,

Nicolle³

et al. 2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

show abstract

“…The pavement images can provide information on the presence of cracks through the pixel intensities and the shape of the darker image features. Many image processing techniques exist for the detection of cracking on grey level images, e.g., [4][5][6][7][8][9][10][11][12][13][14].…”

Section: Introductionmentioning

confidence: 99%

“…The benchmarking of some existing techniques is provided in [1,[4][5][6][7]. One of the latest technique, namely the Minimal Path Selection (MPS) technique has shown to outperform the other methods at the pixel scale on simulated and field pavement images.…”

Section: Introductionmentioning

confidence: 99%

Ongoing tests and improvements of the MPS algorithm for the automatic crack detection within grey level pavement images

Baltazart¹,

Nicolle²,

Yang

2017

2017 25th European Signal Processing Conference (EUSIPCO)

View full text Add to dashboard Cite

Abstract-The MPS approach (Minimal Path Selection) has shown in [1] to provide robust and accurate segmentation of cracks within pavement images compared to other algorithms. As a counterpart, MPS suffers from a large computing time. In this paper, we present three different ongoing improvements to reduce the computing time and to improve the overall segmentation performance. Most of the work focuses on the first three steps of the algorithm which achieve the segmentation of the crack skeleton. This is at first the improvement of the MPS methodology under Matlab coding, then, the C language MPS version and finally, the first attempt to parallelize MPS under the GPU platform. The results on pavement images illustrate the achieved improvements in terms of better segmentation and faster computational time.

show abstract

Crack segmentation in the wild using convolutional neural networks and bootstrapping

Ahmad,

Gharehbaghi,

et al. 2023

Earthquake Engineering and Resilience

View full text Add to dashboard Cite

Computer vision and deep learning methods have found numerous practical applications over the last decade. The field of structural health monitoring has greatly benefited from such advancements. Accurate crack detection and segmentation are a critical part of structural health monitoring and assessment. In the past decade, researchers have developed different computer vision and deep learning methods to address this challenging task. In this article, we propose to use convolutional neural networks and bootstrapping to improve crack segmentation in the wild. Specifically, we investigate fully convolutional network (FCN) and DeepLabV3 with different ResNet architectures as the back bone and assess their performance. A unique feature of this work is the use of bootstrapping with a segmentation network and image augmentation. Bootstrapping is an important component of our proposed methodology for better extracting features from the datasets. We also incorporate superpixel pooling for FCN which improves performance. We assess the performance of our method using five publicly available datasets, which include a wide variety of crack images such as thin pavement cracks, tree shadows on cracks, cracks in asphalt roads, concrete wall cracks, wide cracks on concrete surfaces, and so on. A comparative study is also performed with several established methods in the literature. Results indicate that overall, our proposed method outperforms other state‐of‐the‐art methods for crack segmentation.

show abstract

Automatic Road Crack Detection Using Random Structured Forests

Cited by 985 publications

References 39 publications

Pseudo-ground truth data collection on pavement images

Pseudo-ground truth data collection on pavement images

Ongoing tests and improvements of the MPS algorithm for the automatic crack detection within grey level pavement images

Crack segmentation in the wild using convolutional neural networks and bootstrapping

Contact Info

Product

Resources

About