M.E. Munich and P. Perona. Visual input for pen based computers. IEEE Trans. Pattern Analysis and Machine Intelligence, 24(3):313-328, 2001.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....such as septambic keyer developed by Mann [4] and Twiddler [5] are interesting new approaches to enter data to wearable computers. Computer vision based man machine communication systems can be developed by taking advantage of the character recognition systems developed in document analysis [6,7,12]. For example, unistroke isolated character recognition systems are successfully used in personal digital assistants in which people feel easier to write rather than type on a small size keyboard [8,9] In addition, human like capabilities such as perception would be a good feature of systems ....

....is the Graffiti . The resulting character recognition system can be also used in mobile communication and computing devices such as mobile phones, laptop computers, handheld computers, and PDAs. The advantages of our computer vision based text entry system compared to other vision based systems [10 12] are the following: The background is controlled by the forearm of the user. Furthermore, if the user wears a unicolor fabric then the tip of the finger or the beam of the pointer can be detected in each image of the video by a simple image processing operation such as thresholding. It is ....

M.E. Munich and P. Perona, Visual input for pen-based computers, #3 th Int. Conf. Pattern Recognition, pp.33-37, Vienna, 1996.

....Technology 33594 Bielefeld, Germany E mail: gernot techfak.uni bielefeld.de Abstract The use of handwriting provides a natural way of interacting with small portable computers. However, in order to capture handwritten text on line special input devices are necessary. Therefore, Munich Perona [13] proposed to use visual input for pen based computers. Writing can then be performed on ordinary paper and pen trajectories are automatically extracted from image sequences recorded during the writing process. On the basis of this work we developed a complete video based on line handwriting ....

....requirements restrict the use of handwriting in human computer interaction there has been little research in alternative methods. One idea first envisioned by Munich Perona in 1996 is to produce handwritten text with an ordinary pen on paper and observe the writing process using a video camera [13]. In principle, the 1 This work was supported by the German Research Foundation (DFG) within project Fi799 1. trajectory of the pen can then be recovered automatically by analyzing the recorded image sequence. In contrast to traditional approaches dealing with on line handwriting data the ....

[Article contains additional citation context not shown here]

M. E. Munich and P. Perona. Visual input for pen-based computers. In Proc. Int. Conf. on Pattern Recognition, volume 3, pages 33--37, Vienna, Austria, 1996.

....As our approach to video based on line handwriting recognition is based on the work by Munich Perona we use the same techniques for pen tracking with slight modifications. In the original study the feasibility of the video based tracking was demonstrated. However, even in more recent work [12] handwriting recognition based on video input was not investigated. A comparable approach using video based data acquisition was recently proposed in [1] Difference images are used to obtain the ink trace left on the paper during writing. Figure 1. Pen tracking: The search window for the ....

M. E. Munich. Visual Input for Pen-based Computers. PhD thesis, California Institute of Technology, Pasadena, California, Jan. 2000.

....with small portable computing devices such as personal digital assistants is the use of handwriting. However, for data acquisition touch sensitive pads, which are limited in size, and special pens are required. In order to render this communication method more natural Munich Perona [11] proposed to visually observe the writing process on ordinary paper and to automatically recover the pen trajectory from video image sequences. On the basis of this work we developed a complete handwriting recognition system based on visual input. In this paper we will describe the methods ....

....(PDAs) also the size of the input field is severely restricted. Therefore, a promising alternative might be to retain ordinary pen and paper to produce handwritten text and capture the writing process by observing it using a video camera an idea first envisioned by Munich Perona in 1996 [11]. In contrast to traditional methods to obtain so called on line handwriting data i.e. the dynamically sampled trajectory of the pen an approach based on visual observation will be faced with severe difficulties in obtaining the trajectory data. As a consequence the input data of the ....

[Article contains additional citation context not shown here]

Mario E. Munich and Pietro Perona. Visual input for pen-based computers. In Proc. Int. Conf. on Pattern Recognition, volume 3, pages 33--37, Vienna, Austria, 1996.

....thus be produced with an ordinary pen on paper. The writing process is observed by a video camera and the pen trajectory is automatically extracted from the recorded image sequence. In the original work the feasibility of the video based tracking was demonstrated. However, even in more recent work [10] handwriting recognition based on video input was not investigated. 1 For an overview of the field of handwriting and character recognition see e.g. 2] A comparable approach using video based data acquisition was recently proposed in [1] Difference images are used to obtain the ink trace ....

Mario E. Munich. Visual Input for Pen-based Computers. PhD thesis, California Institute of Technology, Pasadena, California, January 2000.

No context found.

M.E. Munich and P. Perona. Visual input for pen based computers. IEEE Trans. Pattern Analysis and Machine Intelligence, 24(3):313-328, 2001.

No context found.

M.E. Munich and P. Perona. Visual input for pen-based computers. In Proc. 13Int. Conf. Pattern Recognition, 1996.

No context found.

Mario E. Munich. Visual Input for Pen-based Computers. PhD thesis, California Institute mariomu/thesis/thesis.ps.gz.

No context found.

M.E. Munich and P. Perona. Visual input for pen-based computers. In Proc. 13

No context found.

Mario E. Munich. Visual Input for Pen-based Computers. PhD thesis, California Institute of Technology, Pasadena, California, January 2000.

....scanners for off line conversion (see [10] However, all of them are bulky, increasing the size and complexity of the whole system. This paper will present a visual interface that can be built using video technology and computer vision techniques. Some related work can be found in references [7, 5] and our work can also be integrated in the so called Digital Desk [14, 13] being developed at XEROX PARC Cambridge Laboratory. Out input system will consist of a camera, a common piece of paper and a common pen (and of course, a computer to do all the calculations) The camera, focused on the ....

M. Munich and P.Perona. Visual input for pen based computers. CNS Technical Report CNS-TR95 -01, California Institute of Technology, 1995.

....video technology and computer vision techniques in order to capture signatures to be used for personal identification. This vision based personal identification system could be integrated as a component of a complete visual pen based computer environment. Some related work can be found in [3] [11]. Handwriting recognition is still an open problem, even though it has been extensively studied for many years. Signature verification is a reduced problem that still poses a real challenge for researchers. The literature on signature verification is quite extensive (see [1] 9] 16] for very ....

....familiar to the system. 3) An unknown pen is used. The familiar pen case is easy to handle: the system may use a previously stored template representing the pen tip and detect its position in the image by correlation. There are a number of methods to initialize the system when the pen is unknown [11]. Tracking the Pen. The second block of the system has the task of finding the position of the pen tip in the current frame of the sequence. The solution of this task is well known in the optimal signal detection literature. The optimal detector is a filter matched to the signal (in our case a ....

[Article contains additional citation context not shown here]

M.E. Munich and P.Perona. Visual input for pen based computers. CNS Technical Report CNS-TR-95-01, California Institute of Technology, 1995.

....in the warping plane that corresponds to pairs of points in the signatures that have the same displacement vectors, providing zero cost in the case in which one signature is a translated version of the other. 3 Experiments The performance of our tracking system has been presented in references [10], 11] In this paper, we focus our experiments on the results of the automatic personal identification. In section 3.1 we present the performance of the time warping algorithm tested on a database of signatures collected with a tablet digitizer. In section 3.2 we show the results of the visual ....

....3.1, right after the signatures have been captured. The training time is variable, depending on the duration of the signatures, experimentally we observe a maximum training time of 5 seconds. We should point out that our visual tracker is working in real time at 30Hz. As shown in our previous work [10], we are able to track the pen tip in conditions of normal cursive or printed handwriting and drawings. However, in the case of Table 1. Performance of the algorithm tested on the database of signatures collected with the tablet. Signer ID Equal error condition FAR 1 condition # FRs # FAs FRR ....

M. Munich and P. Perona. Visual input for pen-based computers. In Proc. 13 th Int. Conf. Pattern Recognition, 1996.

....area of the parallelogram defined by the vectors C s and 2 C s 2 is constant. To re parameterize the curve we use the following s(p) R p 0 fi fi C t Theta 2 C t 2 fi fi 1 3 dt. 3 Experiments The performance of the visual tracking system has been presented in reference [10] and some preliminary results on the topic of this paper have been described in reference [11] We focus our experiments on the evaluation of the performance of the visually based automatic personal identification system using different parameterization of the signature, enhancing the number of ....

....real ones as a function of the classification threshold. The test set al..lows us to compute the FRR. We computed the FAR in two different ways. First, we used all the signatures from the other subjects as random forgeries, and second, we used the acquired forgeries. As shown in our previous work [10], we are able to track the pen tip in conditions of normal cursive or printed handwriting and drawings. However, in the case of signatures, we observe that the system occasionally loses track of the pen tip when the subject produces an extremely fast stroke. This problem of losing track of the pen ....

M.E. Munich and P. Perona. Visual input for pen-based computers. In Proc. 13 th Int. Conf. Pattern Recognition, 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC