The Experts below are selected from a list of 294 Experts worldwide ranked by ideXlab platform
Enhua Wu - One of the best experts on this subject based on the ideXlab platform.
-
Constant Time Weighted Median Filtering for Stereo Matching and Beyond
2013 IEEE International Conference on Computer Vision, 2013Co-Authors: Kaiming He, Enhua WuAbstract:Despite the continuous advances in local stereo matching for years, most efforts are on developing robust cost computation and aggregation methods. Little attention has been seriously paid to the disparity refinement. In this work, we study weighted median Filtering for disparity refinement. We discover that with this refinement, even the simple Box Filter aggregation achieves comparable accuracy with various sophisticated aggregation methods (with the same refinement). This is due to the nice weighted median Filtering properties of removing outlier error while respecting edges/structures. This reveals that the previously overlooked refinement can be at least as crucial as aggregation. We also develop the first constant time algorithm for the previously time-consuming weighted median Filter. This makes the simple combination ``Box aggregation + weighted median'' an attractive solution in practice for both speed and accuracy. As a byproduct, the fast weighted median Filtering unleashes its potential in other applications that were hampered by high complexities. We show its superiority in various applications such as depth up sampling, clip-art JPEG artifact removal, and image stylization.
Jiyeon Choi - One of the best experts on this subject based on the ideXlab platform.
-
Selection of cost-effective Green Stormwater Infrastructure (GSI) applicable in highly impervious urban catchments
KSCE Journal of Civil Engineering, 2018Co-Authors: Jiyeon Choi, Marla C. Maniquiz-redillas, Jungsun HongAbstract:Urban areas such as roads, and parking lots, present different sets of problem, vehicular Nonpoint source (NPS) loadings derived from vehicular use are very high, while the roof have low NPS pollutant loadings. Therefore, this study was conducted to monitor the actual stormwater runoff characteristics of various landuse types and the treatment efficiency of the constructed GSI systems applied. The data used to calculate the pollutant concentrations were gathered from a total of 172 storm events during the five year monitoring period on a paved road, parking lot and roof landuse site. Based on the results, the road and parking lot landuses have characteristics of large amount of stormwater runoff, high peak flow and runoff of high pollutant concentration due to the vehicular activities. Applicable facilities include pretreatment facilities, such as infiltration trench, subsurface flow (SSF) and hybrid constructed wetland and tree Box Filter which have SA/CA ratios within 1∼2% were appropriate for facilities effective for reducing pollutants including infiltration and filtration functions. Meanwhile, the roof landuse contains low pollutants in comparison to other land uses, so, bioretention, rain garden, and free water surface (FWS) constructed wetland which have SA/CA ratio within 5% were appropriate to enable processing and recycling of large amount of stormwater runoff that have infiltration and retention function. Therefore, costeffective GSI design must not only depend on treated runoff quantity but also quality of the treated runoff for landuse.
-
development of tree Box Filter lid system for treating road runoff
Journal of Wetlands Research, 2013Co-Authors: Jiyeon ChoiAbstract:Abstract The aim of this study was to develop a tree Box Filter system, an example of Low Impact Development technology, for treating stormwater runoff from road. Monitoring of storm events was performed between June 2011 and November 2012 to evaluate the system performance during wet day. Based on the results, all runoff volume generated by rainfall less than 2 mm was stored in the system. The minimum volume reduction of 20% was observed in the system for rainfall greater than 20 mm. The greatest removal efficiency was exhibited by the system for total heavy metals ranging from 70 to 73% while satisfactory removal efficiency was exhibited by the system for particulate matters, organic matters and nutrients ranging from 60 to 68%. The system showed greater pollutant removal efficiency of 67 to 83% for rainfall less than 10 mm compared to rainfall greater than 10 mm which has 39 to 75% pollutant removal efficiency. The system exhibited less pollutant reduction for rainfall greater than 10 mm due to the decreased retention capacity of the system for increased rainfall. Overall, the system has proved to be an option for stormwater management that can be recommended for on-site application. Similar system may be designed based on several factors such as rainfall depth, facility size and pollutant removal efficiency.
Kaiming He - One of the best experts on this subject based on the ideXlab platform.
-
Constant Time Weighted Median Filtering for Stereo Matching and Beyond
2013 IEEE International Conference on Computer Vision, 2013Co-Authors: Kaiming He, Enhua WuAbstract:Despite the continuous advances in local stereo matching for years, most efforts are on developing robust cost computation and aggregation methods. Little attention has been seriously paid to the disparity refinement. In this work, we study weighted median Filtering for disparity refinement. We discover that with this refinement, even the simple Box Filter aggregation achieves comparable accuracy with various sophisticated aggregation methods (with the same refinement). This is due to the nice weighted median Filtering properties of removing outlier error while respecting edges/structures. This reveals that the previously overlooked refinement can be at least as crucial as aggregation. We also develop the first constant time algorithm for the previously time-consuming weighted median Filter. This makes the simple combination ``Box aggregation + weighted median'' an attractive solution in practice for both speed and accuracy. As a byproduct, the fast weighted median Filtering unleashes its potential in other applications that were hampered by high complexities. We show its superiority in various applications such as depth up sampling, clip-art JPEG artifact removal, and image stylization.
Nanning Zheng - One of the best experts on this subject based on the ideXlab platform.
-
sWMF: Separable weighted median Filter for efficient large-disparity stereo matching
2017 IEEE International Symposium on Circuits and Systems (ISCAS), 2017Co-Authors: Shiqiang Chen, Xuchong Zhang, Nanning ZhengAbstract:Although large disparity stereo matching is critical to the practical application of stereo vision system especially for outdoor scenes, its efficient hardware design is still a grand challenge. Motivated by the discovery that well-designed weighted median Filter (WMF) can achieve satisfactory accuracy with simple Box-Filter aggregation, this paper proposes a separable weighted median Filter (sWMF) that only has the computational complexity of O(r) and is independent of disparity range. Moreover, the proposed sWMF can be efficiently implemented as a fully pipelined architecture. Evaluation results demonstrate that, at the penalty of only 0.06% disparity error rate, the proposed sWMF design can save 12.9% Slice LUTs, 76.7% DSPs and 64.0% Block RAMs at the disparity range of 128, compared with previous WMF implementation on FPGA.
Matthew Turk - One of the best experts on this subject based on the ideXlab platform.
-
Cascade of Box (CABox) Filters for Optimal Scale Space Approximation
2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2014Co-Authors: Victor Fragoso, Gaurav Srivastava, Abhishek Nagar, Zhu Li, Kyungmo Park, Matthew TurkAbstract:Local image features, such as blobs and corners, have proven to be very useful for several computer vision applications. However, for enabling applications such as visual search and augmented reality with near-realtime latency, blob detection can be quite computationally expensive due to numerous convolution operations. In this paper, we present a sparse convex formulation to determine a minimal set of Box Filters for fast yet robust approximation to the Gaussian kernels used for blob detection. We call our feature detector as CABox (CAscade of Box) detector. Although Box approximations to a Filter have been studied in the literature, previous approaches suffer from one or more of the following problems: 1) ad hoc Box Filter design, 2) non-elegant trade-off between Filter reconstruction quality and speed and, 3) limited experimental evaluation considering very small datasets. This paper, on the other hand, contributes: 1) an elegant optimization approach to determine an optimal sparse set of Box Filters, and 2) a comprehensive experimental evaluation including a large scale image matching experiment with about 16 K matching and 170 K non-matching image pairs. Our experimental results show a substantial overlap (89%) between the features detected with our proposed method and the popular Difference-of-Gaussian (DoG) approach. And yet CABox is 44% faster. Moreover, the large scale experiment shows that CABox closely reproduces DoG's performance in an end-to-end feature detection and matching pipeline.
-
CVPR Workshops - Cascade of Box (CABox) Filters for Optimal Scale Space Approximation
2014 IEEE Conference on Computer Vision and Pattern Recognition Workshops, 2014Co-Authors: Victor Fragoso, Gaurav Srivastava, Abhishek Nagar, Zhu Li, Kyungmo Park, Matthew TurkAbstract:Local image features, such as blobs and corners, have proven to be very useful for several computer vision applications. However, for enabling applications such as visual search and augmented reality with near-realtime latency, blob detection can be quite computationally expensive due to numerous convolution operations. In this paper, we present a sparse convex formulation to determine a minimal set of Box Filters for fast yet robust approximation to the Gaussian kernels used for blob detection. We call our feature detector as CABox (CAscade of Box) detector. Although Box approximations to a Filter have been studied in the literature, previous approaches suffer from one or more of the following problems: 1) ad hoc Box Filter design, 2) non-elegant trade-off between Filter reconstruction quality and speed and, 3) limited experimental evaluation considering very small datasets. This paper, on the other hand, contributes: 1) an elegant optimization approach to determine an optimal sparse set of Box Filters, and 2) a comprehensive experimental evaluation including a large scale image matching experiment with about 16 K matching and 170 K non-matching image pairs. Our experimental results show a substantial overlap (89%) between the features detected with our proposed method and the popular Difference-of-Gaussian (DoG) approach. And yet CABox is 44% faster. Moreover, the large scale experiment shows that CABox closely reproduces DoG's performance in an end-to-end feature detection and matching pipeline.
