imageprocessingalgorithms used in surveillance systems are designed to work under good weather conditions. For example, in a rainy day, raindrops are adhered to camera lenses and windshields, resulting in partial oc...
详细信息
ISBN:
(纸本)9789881476807
imageprocessingalgorithms used in surveillance systems are designed to work under good weather conditions. For example, in a rainy day, raindrops are adhered to camera lenses and windshields, resulting in partial occlusions in acquired images, and making performance of imageprocessingalgorithms significantly degraded. To improve performance of surveillance systems in a rainy day, raindrops have to be automatically detected and removed from images. Addressing this problem, this paper proposes an adherent raindrop detection method from a single image which does not need training data and special devices. The proposed method employs image segmentation using Maximally Stable Extremal Regions (MSER) and qualitative metrics to detect adherent raindrops from the result of MSER-based image segmentation. Through a set of experiments, we demonstrate that the proposed method exhibits efficient performance of adherent raindrop detection compared with conventional methods.
The exponentially enlarging usability and mushrooming features in the portable hand-held devices such as cellphone appeal for the highly energy-efficient designs as the user cannot tolerate power famished devices rath...
详细信息
The exponentially enlarging usability and mushrooming features in the portable hand-held devices such as cellphone appeal for the highly energy-efficient designs as the user cannot tolerate power famished devices rather can manage with tiny degraded quality. The multimedia applications such as image/video processing are compute-intensive and require efficient designs which can be achieved by approximate designs with the feeble accord in accuracy. As multiplication is the most frequently used and highly energy-consuming arithmetic operation, in this paper, a low power approximate multiplier architecture (AMA) is proposed that significantly improves the design metrics. We evaluate the proposed AMA with various quality metrics and design an AMA embedded Gaussian smoothing filter (GSF). The simulation results show that AMA reduces 81.5% power consumption over accurate and 19.17% energy over the well-known approximate multiplier architecture. Further, the AMA embedded GSF reduces 6.2% area and 7.53% power consumption over the best known approximate multiplier embedded GSF.
This paper presents vehicle black box using camera image and 24GHz Frequency Modulation Continuous Wave (FMCW) radar. Currently, almost all of vehicle black boxes are recording driving data using a single camera. They...
详细信息
ISBN:
(纸本)9781509025985
This paper presents vehicle black box using camera image and 24GHz Frequency Modulation Continuous Wave (FMCW) radar. Currently, almost all of vehicle black boxes are recording driving data using a single camera. They have some problems on low image quality and narrow viewing angle. In addition, it is difficult to record a clean image when the weather is in bad condition due to heavy rain, snow and night. These problems may have some troubles to investigate car accident accurately. In this paper, we propose a novel black box fusing data from radar and cameras. We estimate the distance and velocity of obstacle using the Doppler Effect and FMCW radar. In particular, we compare various Direction of Arrival (DOA) algorithms for estimation of obstacle angle. The experimental results indicate that the Multiple Signal Classification (MUSIC) is more accurate compared to other methods. Moreover, we formulate the vehicle tracking image through the camera imageprocessing and radar information. This image can be used to widen viewing angle and it may be helpful to investigate a car accident. The experimental study results indicate that the new black box overcome the disadvantages of the current black boxes and may be useful for safe driving and accurate accident investigation.
Robotic vision, unlike computer vision, typically involves processing a stream of images from a camera with time varying pose operating in an environment with time varying lighting conditions and moving objects. Repea...
详细信息
ISBN:
(纸本)9781509037636
Robotic vision, unlike computer vision, typically involves processing a stream of images from a camera with time varying pose operating in an environment with time varying lighting conditions and moving objects. Repeating robotic vision experiments under identical conditions is often impossible, making it difficult to compare different algorithms. For machine learning applications a critical bottleneck is the limited amount of real world image data that can be captured and labelled for both training and testing purposes. In this paper we investigate the use of a photo-realistic simulation tool to address these challenges, in three specific domains: robust place recognition, visual SLAM and object recognition. For the first two problems we generate images from a complex 3D environment with systematically varying camera paths, camera viewpoints and lighting conditions. For the first time we are able to systematically characterise the performance of these algorithms as paths and lighting conditions change. In particular, we are able to systematically generate varying camera viewpoint datasets that would be difficult or impossible to generate in the real world. We also compare algorithm results for a camera in a real environment and a simulated camera in a simulation model of that real environment. Finally, for the object recognition domain, we generate labelled image data and characterise the viewpoint dependency of a current convolution neural network in performing object recognition. Together these results provide a multi-domain demonstration of the beneficial properties of using simulation to characterise and analyse a wide range of robotic vision algorithms.
This paper proposes a vehicle detection algorithm using pre-processing and lamp detection at night-time. First, we present a vehicle detection using contrast enhancement. By applying a specific contrast enhancement to...
详细信息
ISBN:
(纸本)9781467380171
This paper proposes a vehicle detection algorithm using pre-processing and lamp detection at night-time. First, we present a vehicle detection using contrast enhancement. By applying a specific contrast enhancement to night-time images with low exposure, we can enhance salient features even at dark night-time. Next, we detect a pair of rear lamps from the pre-processed image(s). Finally, we can find forward vehicle(s) by lamp pairing. Experimental results show that the proposed algorithm provides reliable detection accuracy.
Word segmentation of handwritten documents is a vital step in the Optical Character Recognition system as its accuracy greatly influences the overall recognition performance. In the literature, various methods have be...
详细信息
ISBN:
(纸本)9781509046218
Word segmentation of handwritten documents is a vital step in the Optical Character Recognition system as its accuracy greatly influences the overall recognition performance. In the literature, various methods have been proposed for word segmentation of handwritten documents of various languages. However, it is observed that for Odia, which is an important Indian language, very little work has been reported on word segmentation. Hence, the objective of this paper is to employ two standard existing methods to segment words of Odia handwritten documents and compare the segmentation performance of these methods with the lone Water Reservoir Algorithm available in the literature and finally rank those methods based on their segmentation performance. It is observed that out of three methods, the Tree Structure method performs the best comparing four different performance measures.
Feature selection, as a preprocessing step to machine learning, plays a pivotal role in removing irrelevant data, reducing dimensionality and improving performance evaluations. Recent years, sparse representation has ...
详细信息
ISBN:
(纸本)9781509055227
Feature selection, as a preprocessing step to machine learning, plays a pivotal role in removing irrelevant data, reducing dimensionality and improving performance evaluations. Recent years, sparse representation has become a useful tool for both supervised and unsupervised feature selection. So far, most of these algorithms still have many problems such as large computation load, performance with poor stability. Thus, this paper proposes a new unsupervised feature selection algorithm via sparse representation (UFSSR), with respect to efficiency and effectiveness. Firstly, this paper reconstructs part of data matrix via sparse representation, which makes the proposed algorithm be robust and independent of domain knowledge. Then, to reduce the reconstruction error, a new feature evaluation function is given to rank all features. Theoretical analysis and experiments compared with many popular algorithms on a set of datasets demonstrate the improvements brought by UFSSR.
image segmentation is a key component in many computer vision systems, and it is recovering a prominent spot in the literature as methods improve and overcome their limitations. The outputs of most recent algorithms a...
详细信息
ISBN:
(纸本)9781467388528
image segmentation is a key component in many computer vision systems, and it is recovering a prominent spot in the literature as methods improve and overcome their limitations. The outputs of most recent algorithms are in the form of a hierarchical segmentation, which provides segmentation at different scales in a single tree-like structure. Commonly, these hierarchical methods start from some low-level features, and are not aware of the scale information of the different regions in them. As such, one might need to work on many different levels of the hierarchy to find the objects in the scene. This work tries to modify the existing hierarchical algorithm by improving their alignment, that is, by trying to modify the depth of the regions in the tree to better couple depth and scale. To do so, we first train a regressor to predict the scale of regions using mid-level features. We then define the anchor slice as the set of regions that better balance between over-segmentation and under-segmentation. The output of our method is an improved hierarchy, re-aligned by the anchor slice. To demonstrate the power of our method, we perform comprehensive experiments, which show that our method, as a post-processing step, can significantly improve the quality of the hierarchical segmentation representations, and ease the usage of hierarchical image segmentation to high-level vision tasks such as object segmentation. We also prove that the improvement generalizes well across different algorithms and datasets, with a low computational cost.
We address the following question: is it possible to reconstruct the geometry of an unknown environment using sparse and incomplete depth measurements? This problem is relevant for a resource-constrained robot that ha...
详细信息
ISBN:
(纸本)9781509037636
We address the following question: is it possible to reconstruct the geometry of an unknown environment using sparse and incomplete depth measurements? This problem is relevant for a resource-constrained robot that has to navigate and map an environment, but does not have enough on-board power or payload to carry a traditional depth sensor (e.g., a 3D lidar) and can only acquire few (point-wise) depth measurements. In general, reconstruction from incomplete data is not possible, but when the robot operates in man-made environments, the depth exhibits some regularity (e.g., many planar surfaces with few edges); we leverage this regularity to infer depth from incomplete measurements. Our formulation bridges robotic perception with the compressive sensing literature in signal processing. We exploit this connection to provide formal results on exact depth recovery in 2D and 3D problems. Taking advantage of our specific sensing modality, we also prove novel and more powerful results to completely characterize the geometry of the signals that we can reconstruct. Our results directly translate to practical algorithms for depth reconstruction; these algorithms are simple (they reduce to solving a linear program), and robust to noise. We test our algorithms on real and simulated data, and show that they enable accurate depth reconstruction from a handful of measurements, and perform well even when the assumption of structured environment is violated.
Many recently proposed graph-processing frameworks utilize powerful computer clusters with dozens of cores to process massive graphs. Their usability and flexibility often come at a cost. We demonstrate that custom so...
详细信息
ISBN:
(纸本)9781467390064
Many recently proposed graph-processing frameworks utilize powerful computer clusters with dozens of cores to process massive graphs. Their usability and flexibility often come at a cost. We demonstrate that custom software written for “nanocomputers,” including a credit-card-sized Raspberry Pi, a low-cost ARM server, and an Intel Atom computer, can process the same graphs. Our implementations of PageRank and connected components stream graphs from external storage while performing computation in the limited main memory on these nanocomputers. The results show that a $100 computer with an Intel Atom core can compute PageRank and connected components on a 1.5-billion-edge Twitter graph as quickly as graph-processingsystems running on machines with up to 48 cores. As people continue to apply graph computations to large datasets, this research suggests that there may be cost and energy advantages to using nanocomputers.
暂无评论