Current Issue : April - June Volume : 2018 Issue Number : 2 Articles : 5 Articles
Multi-modal sensory data plays an important role in many computer vision and robotics tasks. One popular\nmulti-modal pair is cameras and laser scanners. To overlay and jointly use the data from both modalities, it is necessary\nto calibrate the sensors, i.e., to obtain the spatial relation between the sensors.\nComputing such a calibration is challenging as both sensors provide quite different data: cameras yield color or\nbrightness information, laser scanners yield 3-D points. However, several laser scanners additionally provide\nreflectances, which turn out to make calibration to a camera well feasible. To this end, we first estimate a rough\nalignment of the coordinate systems of both modalities. Then, we use the laser scanner reflectances to compute a\nvirtual image of the scene. Stereo calibration on the virtual image and the camera image are then used to compute a\nrefined, high-accuracy calibration.\nIt is encouraging that the accuracies in our experiments are comparable to camera-camera stereo setups and\noutperform another of other target-based calibration approach. This shows that the proposed algorithm reliably\nintegrates the point cloud with the intensity image. As an example application, we use the calibration results to\nobtain ground-truth distance images for range cameras. Furthermore, we utilize this data to investigate the accuracy\nof the Microsoft Kinect V2 time-of-flight and the Intel RealSense R200 structured light camera....
The objective of this study is to demonstrate through empirical evaluation the potential of a number of computer\nvision (CV) methods for sex determination from human skull. To achieve this, six local feature representations, two\nfeature learnings, and three classification algorithms are rigorously combined and evaluated on skull regions derived\nfrom skull partitions. Furthermore, we introduce for the first time the application of multi-kernel learning (MKL) on\nmultiple features for sex prediction from human skull. In comparison to the classical forensic methods, the results in\nthis study are competitive, attesting to the suitability of CV methods for sex estimation. The proposed approach is fully\nautomatic....
In sports analysis, player tracking is essential to the extraction of statistics such as speed, distance and direction of\nmotion. Simultaneous tracking of multiple people is still a very challenging computer vision problem to which there is\nno satisfactory solution. This is especially true for sports activities, for which people often wear similar uniforms, move\nquickly and erratically, and have close interactions with each other. In this paper, we introduce a multi-target tracking\nalgorithm suitable for team sports activities. We extend an existing algorithm by including an automatic estimation of\nthe occupancy of the observed field and the duration of stable periods without people entering or leaving the field.\nThis information is included as a constraint to the existing offline tracking algorithm in order to construct more reliable\ntrajectories. On data from two challenging sports scenariosââ?¬â?an indoor soccer game captured with thermal cameras\nand an outdoor soccer training session captured with RGB cameraââ?¬â?we show that the tracking performance is\nimproved on all sequences. Compared to the original offline tracking algorithm, we obtain improvements of 3ââ?¬â??7% in\naccuracy. Furthermore, the method outperforms two state-of-the-art trackers....
In-field mango fruit sizing is useful for estimation of fruit maturation and size distribution,\ninforming the decision to harvest, harvest resourcing (e.g., tray insert sizes), and marketing. In-field\nmachine vision imaging has been used for fruit count, but assessment of fruit size from images\nalso requires estimation of camera-to-fruit distance. Low cost examples of three technologies for\nassessment of camera to fruit distance were assessed: a RGB-D (depth) camera, a stereo vision camera\nand a Time of Flight (ToF) laser rangefinder. The RGB-D camera was recommended on cost and\nperformance, although it functioned poorly in direct sunlight. The RGB-D camera was calibrated,\nand depth information matched to the RGB image. To detect fruit, a cascade detection with histogram\nof oriented gradients (HOG) feature was used, then Otsu�s method, followed by color thresholding\nwas applied in the CIE L*a*b* color space to remove background objects (leaves, branches etc.).\nA one-dimensional (1D) filter was developed to remove the fruit pedicles, and an ellipse fitting\nmethod employed to identify well-separated fruit. Finally, fruit lineal dimensions were calculated\nusing the RGB-D depth information, fruit image size and the thin lens formula. A Root Mean Square\nError (RMSE) = 4.9 and 4.3 mm was achieved for estimated fruit length and width, respectively,\nrelative to manual measurement, for which repeated human measures were characterized by a\nstandard deviation of 1.2 mm. In conclusion, the RGB-D method for rapid in-field mango fruit size\nestimation is practical in terms of cost and ease of use, but cannot be used in direct intense sunshine.\nWe believe this work represents the first practical implementation of machine vision fruit sizing in\nfield, with practicality gauged in terms of cost and simplicity of operation....
The detection of rail surface defects is an important part of railway daily inspection, according to the requirements\nof modern railway automatic detection technology on real-time detection and adaptability. This paper presents a\nmethod for real-time detection of rail surface defects based on machine vision. According to the basic principle of\nmachine vision, an image acquisition device equipped with LED auxiliary light source and shading box has been\ndesigned and the portable testing model is designed to carry on the field experiment. In view of the real-time\nrequirement, the method of extracting the target area from the original image is carried out without image preprocessing.\nThe surface defects of the rail are optimized based on morphological process and the characteristics of\nthe defects are obtained by tracking the direction chain code. It is demonstrated that the maximum positioning\ntime of this proposed method is 4.65 ms and its maximum positioning failure rate is 5%. The real-time detection\nspeed of this proposed method can reach 2 m/s, which can carry out real-time detection of artificial hand walking.\nThe time of processing each picture is up to 245.61 ms, which ensures the real-time performance of the portable\ntrack defect vision inspection system. To a certain extent, the system can replace manual inspection and carry out\nthe digital management of track defects....
Loading....