AprilTags and other passive fiducial markers require specialized algorithms to detect markers among other features in a natural scene. The vision processing steps generally dominate the computation time of a tag detec...
详细信息
ISBN:
(纸本)9781509037636
AprilTags and other passive fiducial markers require specialized algorithms to detect markers among other features in a natural scene. The vision processing steps generally dominate the computation time of a tag detection pipeline, so even small improvements in marker detection can translate to a faster tag detection system. We incorporated lessons learned from implementing and supporting the AprilTag system into this improved system. This work describes AprilTag 2, a completely redesigned tag detector that improves robustness and efficiency compared to the original AprilTag system. The tag coding scheme is unchanged, retaining the same robustness to false positives inherent to the coding system. The new detector improves performance with higher detection rates, fewer false positives, and lower computational time. Improved performance on small images allows the use of decimated input images, resulting in dramatic gains in detection speed.
Convolutional neural networks (CNNs), in combination with big data, are increasingly being used to engineer robustness into visual classification systems including human detection. One significant challenge to using a...
详细信息
ISBN:
(纸本)9781509039302
Convolutional neural networks (CNNs), in combination with big data, are increasingly being used to engineer robustness into visual classification systems including human detection. One significant challenge to using a CNN on a mobile robot, however, is the associated computational cost and detection rate of running the network. In this work, we demonstrate how fusion with a feature-based layered classifier can help. Not only does score-level fusion of a CNN with the layered classifier improve precision/recall for detecting people on a mobile robot, but using the layered system as a pre-filter can substantially reduce the computational cost of running a CNN - reducing the number of objects that need to be classified while still improving precision. The combined real-time system is implemented and evaluated on a two robots with very different GPU capabilities.
The concealed weapon, like blade, detection and identification is one of the most puzzling task faces by security agency. Researchers have demonstrated MMW imaging systems to detect concealed targets like gun, knife a...
详细信息
ISBN:
(纸本)9781509033331
The concealed weapon, like blade, detection and identification is one of the most puzzling task faces by security agency. Researchers have demonstrated MMW imaging systems to detect concealed targets like gun, knife and scissors but detection of small size target like blade with different orientation is still challenging due to resolution limitation of MMW imaging system. The success of small size concealed target detection depends upon scanning step size of imaging system and dielectric property of covering cloths and hidden object. Therefore, resolution enhancement techniques may play a very important role for small size concealed target detection. To perceive such challenges, active V-band MMW radar conjunction with imageprocessing techniques has been demonstrated for detection and identification of concealed blade and obtained two dimensional good quality of images of concealed blade under different cloths at various angle. For this purpose, a critical analysis of various signal and imageprocessing has been carried out and integrated following algorithms like singular value decomposition (SVD) for clutter reduction, discrete wavelet transform (DWT) for resolution enhancement, thresholding for target detection and in last artificial neural network (ANN) based algorithm for rotation invariant target identification. An imageprocessing based methodology has been proposed by which the concealed target like blade can be successfully detected.
We propose information processing techniques for CCTV based surveillance systems employed in (a) work environments and (b) public places and transport, for automated identification of scenes of inter-personal crime. A...
详细信息
ISBN:
(纸本)9781467385497
We propose information processing techniques for CCTV based surveillance systems employed in (a) work environments and (b) public places and transport, for automated identification of scenes of inter-personal crime. Although both the scenarios presented in this work employ similar signal processing and learning algorithms, the objective involved are significantly different. In (a) we aim to preserve confidentiality and privacy of official meetings and discussions, while ensuring detection of un-becoming behavior, like: bullying, harassment and assault. In the proposed method we identify such critical conditions using a combination of image and speech processing and ensue conditional video recording and saving. In (b), the target is to identify the occurrence of interpersonal crime using video and voice processing, in order to raise alert at the local surveillance station, which may be receiving numerous CCTV videos from neighboring areas. This can be an assistance to the security personnel, responsible to monitor large number of screens. The proposed methods can be useful curbing inter-personal violence, and crime against women, in the form of eve teasing, and harassment.
Lane detection algorithm using a vision sensor or a camera would be more effective for self-driving vehicles to keep in lane, if it is possible to derive a distance ratio between a vehicle and left-right lanes. Howeve...
详细信息
ISBN:
(纸本)9781509032204
Lane detection algorithm using a vision sensor or a camera would be more effective for self-driving vehicles to keep in lane, if it is possible to derive a distance ratio between a vehicle and left-right lanes. However, a dangerous situation may occur if the performance of the camera (e.g., frame/sec.) and the real-time speed of the vehicle are not considered properly because of the huge distance difference among frames for a fast moving vehicle with a low-speed camera. In this study, we propose a simple method to anticipate the relative position of the vehicle in the following frame from the current frame image. The expected ratio between a vehicle and the left-right lanes can be obtained by using of the speed of a vehicle and the frame speed of a camera. Experiment results show that less than 5.28% error occurs by the proposed algorithm for various cars and cameras.
This paper presents a robust hand detection algorithm using the facial information. The proposed algorithm consists of four steps: (i) detection of a face, (ii) generation of regions of interest (ROI) to detect hands,...
详细信息
This paper presents a robust hand detection algorithm using the facial information. The proposed algorithm consists of four steps: (i) detection of a face, (ii) generation of regions of interest (ROI) to detect hands, (iii) skin color extraction from the face region, and (iv) detection of hands using the face skin color in the ROI. The proposed algorithm can reduce false detection caused by a similar skin color, and provides a successful detection rate up to 92 percent.
Automatic object recognition for texture-less objects using computer vision is a difficult task in comparison of textured one since class discriminative information is rarely available. Herein, an algorithm to count s...
详细信息
Automatic object recognition for texture-less objects using computer vision is a difficult task in comparison of textured one since class discriminative information is rarely available. Herein, an algorithm to count such objects using shape and color attributes for recognition with scale, rotation and illumination invariance is proposed. Initially, the algorithm extracts shape and color features of the prototype image to find its instance in the real-time pre-processed scene image captured by the vision interface. The pre-processing is achieved by morphological boundary extraction and segmentation techniques. Color and shape features are extracted based on mean hue value and Hu-moments respectively from the obtained segments. SVM, kNN, neural network and tree-bagging are then applied for classification. Tree-bagging is found to eclipse over the other classifiers in terms of accuracy. Finally, the classified objects are counted and localized in the image by drawing bounding boxes around them. A desktop application of the proposed algorithm is also developed. To assess the performance of the proposed algorithm, experimentation has been carried out for various objects having different shapes and colors. The algorithm proved out to be robust and effective for recognition and counting of the texture-less objects.
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.
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.
Sparse representation model adopts an image patch as a linear combination of a few atoms chosen out from an overcomplete dictionary, and they have shown promising results in various image restoration applications. Gra...
详细信息
暂无评论