Search

Feature phones or low-end smart phones are an important part of our daily life. While lacking the powerful embedded camera and processing capability of the high-end smart phone, the feature phones with the low-end camera and processing capability are still widely used owing to its lower cost. Consequently, image and video processing algorithms are needed for many intelligent applications on the low-end smart phones. In this paper, we propose a novel robust and efficient video anti-shaking algorithm for the low-end smartphone platform. It provides an integrated application for the low-end smartphone platform, which includes a novel global motion estimation method, and a rapid and efficient local motion detection method. The overall framework is proposed considering the limited camera perception quality and limited mobile processing capability of the feature phone. The proposed algorithm can not only obtain a good visual effect but also be implemented in real-time for the low-end mobile platforms with very limited computational resource. Experimental results demonstrate that the proposed method can achieve a similar video stabilization effect to the Optical Image Stabilization hardware in high-end smartphones.

In this paper, we developed a golf position analysis system with an inertial sensor and swing motion analysis algorithm that can be used in golf courses. Specifically, we developed a band for both wrists that analyzes the golf swing of whoever is wearing it. By analyzing the movements of both the right and left wrists of the golfer, their golf swing can be judged more accurately, which can be used to advise the golfer regarding their golf swing. Experimental results show that this portable instrument along with its swing motion detection algorithm is a promising training tool for golfers.

This paper describes a system of practical technologies to implement an illumination-robust, consumer grade biometric system based on face recognition to be used in the automotive market. Most current face recognition systems are compromised in accuracy by ambient illumination changes. Outdoor applications including vehicle personalization pose an especially challenging environment for face recognition. The point of this research is to investigate practical face recognition used for identity management in order to minimize algorithmic complexity while making the system robust to ambient illumination changes. First, a frame differencing method is presented with an active near-infrared illumination control that produces images independent of the ambient illumination. Second, an end-to-end face recognition system is presented including foreground/background segmentation, motion detection, face detection, motion interpolation, pose clustering and face recognition modules. It is shown that the frame differencing method makes the modules more robust to the ambient illumination variation. Vehicular application videos were taken in extremely challenging outdoor illumination and shadowing conditions and used to test each module. Finally, extensive test results of vehicular scenario are provided to evaluate the end-to-end systems1.

Proximity-based motion gesture sensors (MGSs) using IR LEDs have been widely researched for contactless human-machine interfaces. Conventional MGSs have incorporated discrete photodiodes into a receiver, which prevents them from being adopted for portable handheld devices because of their large size and high profile. This paper proposes a proximity-based contactless MGS that embeds photodiodes and a motion detection algorithm into a single package. The area and height of the proposed optical structure including photodiodes and an optical block are reduced by 96.2% and 71.7%, respectively, compared with the most recently developed MGS. Furthermore, the proposed sensor achieves a recognition rate of over 99.5% on average at various speeds of hand movements. A prototype system is implemented and tested for validation1.

This paper presents a new algorithm for detecting and analyzing the periodic motions in video sequences. Different from the previous methods which detect periodic motions from the entire frame, we propose a convexhull- based process to automatically determine the regions of interest (ROI) of the motions and utilize an ROI-based similarity measure to detect the motion periods. Furthermore, we also propose an extrema-based method to select the optimal reference frame for further improving the periodic detection performance. Our proposed algorithm can not only effectively detect motion periods with both constant and variable period lengths, but also have obvious advantage when handling periodic motion with slight movements. Experimental results demonstrate the effectiveness of our proposed method.

This paper proposes a weighted sum based multi-exposure image fusion method which consists of two main steps: three image features composed of local contrast, brightness and color dissimilarity are first measured to estimate the weight maps refined by recursive filtering. Then, the fused image is constructed by weighted sum of source images. The main advantage of the proposed method lies in a recursive filter based weight map refinement step which is able to obtain accurate weight maps for image fusion. Another advantage is that a novel histogram equalization and median filter based motion detection method is proposed for fusing multi-exposure images in dynamic scenes which contain motion objects. Furthermore, the proposed method is quite fast and thus can be directly used for most consumer cameras. Experimental results demonstrate the superiority of the proposed method in terms of subjective and objective evaluation.

In this paper we present a way to reduce the standby power consumption of a PIR-sensor-based lighting device. Generally although a PIR-sensor-based lighting device will turn on when motion is detected, and will turn off when the motion disappears, the device still consumes 1–3 W power when the lamp is off. In our design the device consumes 0.007 W when the light is off, and is not only easy to set up but also inexpensive. Our circuit supplies the lamp with power when motion is detected; when the motion disappears it turns the lamp off, and the electric power is shut off in order to reduce the standby power. We use an MCU which receives signals from a PIR sensor which detects any individual approaching the device. The MCU controls the SSR On/Off when used as a lighting switch for shutting off the standby power. The MCU monitoring program provides automatic detection of any individual by means of the PIR sensor. The MCU has internal modules to simplify the hardware circuit design. The circuit component count, cost and power consumption are low.

Markerless camera-based games are more immersive and compelling than other types of games. However, it is difficult to select the proper methods for body segmentation and motion detection, because they should be fast processed for real-time responses. Therefore, collision detection and response between virtual objects (VOs) and human bodies are estimated using collision regions. In this paper, we propose a novel collision detection and response method capable of overcoming many disadvantages of previous markerless games. The proposed method is based on frame difference processing, and estimates motion boundaries (MBs); therefore, the method is fast processed. The MBs, which consist of vertices having motion vectors (MVs), are valid boundaries that can detect collisions with VOs. To estimate MVs and MBs, fast polygonal approximation and vertex matching are applied. Thus, MBs can be fast estimated, collisions can be correctly detected, and VOs respond to moving bodies instead of collision regions. The experimental results of the collision detection and response between human and VOs show that the proposed method properly estimates MBs and guarantees precise collision responses. Therefore, the proposed method can be used in camera-based games requiring high performance.

This paper presents an efficient and effective motion blur removal method based on long and short exposure images. The two images are captured sequentially and motion pixels between the images are then robustly detected, with suppression of noise and prevention of artifacts around object boundaries. Object motion blur is removed and high quality image is obtained by merging the two images with taking into account the detected motion pixels. The proposed method is directly performed on the CFA (Color Filter Array) image which only has one color component per pixel. It has low computational complexity and low memory requirements. The proposed method also achieves a HDR (High Dynamic Range) image at the same time.

This paper proposes a CMOS motion detector which consumes extremely low power. CMOS image sensor pixels in this motion detector senses image and image data are converted into just one-bit by using clocked comparators. Because using one-bit data makes additional processing units simple, total power consumption of this CMOS motion detector can be reduced. That is, internal memory which is composed of the clocked gating schemes based on the flip-flop and XOR function which compares a current image with a previous one in order to detect a difference are main features for the simple structure. However one-bit data process has a critical problem that it is hard to detect a motion when image inclines to white or block. For solving this problem, reference voltage controller which makes about the same proportion of white and block is implemented. We have made a test module of the proposed CMOS motion detector and tested it by using FPGA. According to the measurement result, total power consumption is about 32 ?W at 3.3 V. Therefore, this motion detector can be useful for portable battery-operated devices.

In this paper, we present a universal remote control system based on computer vision. The method is composed of two stages of detecting visual evidences. Motion and skin color information is utilized to detect waving hands requesting control commands. Upon the request, the camera is controlled to zoom in on local region of the hand. The number of open fingers is counted by the shape analysis on the segmented hand region image. Control command is issued when a predetermined sequence of gesture state transition is produced. Experimental results show that the shape of open fingers exhibits strong features for determining correct gesture states. The use of stable features on consecutive frames yields robust and accurate performance regardless of operating conditions.

This paper focuses on the rolling shutter distortion of CMOS image sensor coming from its unique readout mechanism as the main cause for image degradation when there are fast-moving objects. This paper proposes a post image processing scheme based on motion vector detection to suppress the rolling shutter distortion. Motion vector detection is performed based on an optical flow method at a reasonable computational complexity. A practical implementation scheme is also described.

In this paper, we propose a new deinterlacing algorithm using edge direction field, edge parity, and motion expansion scheme. The algorithm consists of an EDI (edge dependent interpolation)-based intra-field deinterlacing and inter-field deinterlacing that uses block-based motion detection. Most of the EDI algorithms use pixel-by-pixel or block-by-block distance to estimate the edge direction, which results in many annoying artifacts. We propose the edge direction field, and estimate an interpolation direction using the field and SAD (sum of absolute differences) values. The edge direction field is a set of edge orientations and their gradient magnitudes. The proposed algorithm assumes that a local minimum around the gradient edge field is most probably the true edge direction. Our approach provides good visual results in various kinds of edges (horizontal, narrow and weak). And we propose a new temporal interpolation method based on block motion detection. The algorithm works reliably in scenes which have very fast moving objects and low SNR signals. Experimental results on various data set show that the proposed algorithm works well for the diverse kinds of sequences and reconstructs flicker-free details in the static region1.

We present a surveillance camera architecture using a one-bit motion detection algorithm for portable applications. The conventional surveillance system requires too many components and is too complicated for portable applications. In this paper, we propose a simple surveillance camera architecture with a motion detector that can detect changes of an object within its surroundings for low-power and low-cost applications. The proposed motion-detection algorithm uses one-bit images which are selected using a binary search algorithm that is decided by varying the reference voltage. We verify all the functions of the motion detector through the use of a user interface program on an FPGA board. Experimental results are also presented1.

Extraction of moving objects is an important and fundamental research topic for many digital video applications. This paper addresses an efficient spatio-temporal segmentation scheme to extract moving objects from video sequences. The temporal segmentation yields a temporal mask that indicates moving regions and static regions for each frame. For localization of moving objects, a block-based motion detection method considering a novel feature measure is proposed to detect changed regions. These changed regions are coarse and need accurate spatial compensation. An edge-based morphological dilation method is presented to achieve the anisotropic expansion of the changed regions. Furthermore, to solve the temporarily stopping problem of moving objects, the inertia information of moving objects is considered in the temporal segmentation. The spatial segmentation based on the watershed algorithm is performed to provide homogeneous regions with closed and precise boundaries. It considers the global information to improve the accuracy of the boundaries. To reduce over-segmentation in the watershed segmentation, a novel mean filter is proposed to suppress some minima. A fusion of the spatial and temporal segmentation results produces complete moving objects faithfully. Compared with the reference algorithms, the fusion threshold in our scheme is fixed for different sequences. Experiments on typical sequences have successfully demonstrated the validity of the proposed scheme.

Automated security systems are a useful addition to today's home where safety is an important issue. Vision-based security systems have the advantage of being easy to set up, inexpensive and non-obtrusive. This paper proposes an integrated dual-level vision-based home security system, which consists of two subsystems - a face recognition module and a motion detection module. The primary face recognition module functions as a user authentication device. On an event of a failure in the primary system, the secondary motion detection module acts as a reliable backup to detect human-related motions within certain locations inside the home. Novel algorithms have been proposed for both subsystems. Several experiments and field tests conducted have shown good performance and feasible implementation in both subsystems.

In this paper, we propose a motion adaptive deinterlacing method with texture detection for scan-rate conversion of video data. The basic idea of this method is to classify the missing pixel to four different regions including motion smooth region, motion texture region, static smooth region, and static texture region based on the results of motion detection and texture detection, then four different interpolation methods are used to produce the missing pixel. Extensive simulations conducted for video sequences indicate that the performance of the proposed method is superior to the previous methods based on the direct merging method, bilinear interpolation, ELA, and vertical temporal median method. Moreover, our proposed method requires lower computational complexity and provides better object and subject performance

This paper presents an efficient and robust adaptive deinterlacing technique that consists of three main steps, i.e. motion estimation, motion detection (MD), and adaptive decision-making. Firstly, five successive interlaced fields are used and four unidirectional motion vectors are calculated. Secondly, a probabilistic criterion is introduced to select the most reliable motion vector for each block before MD, which has four possible output motion states. Finally, an adaptive deinterlacing scheme is proposed based on the motion and texture information of each block, as well as the output of MD. Experimental results show that the proposed deinterlacing technique outperforms previous methods both objectively and subjectively. The analysis also shows that the proposed technique is robust to both unconventional motions and the video codec

In this paper a video processing engine with accurate motion detection and sawtooth artifacts remove capabilities for LCD TV is proposed. Since the accuracy of the motion detection is key factor to obtain excellent visual quality pictures, by incorporating the temporal and spatial information of video fields and sophisticated detection schemes, the goal can be obtained. Also, a novel window-expanding search approach for low-angle edge detection/interpolation to remove the sawtooth artifacts, which occur on the edges/lines of fast moving images while doing the de-interlacing, is proposed. To validate the proposed methods and consider the bandwidth requirement, an FPGA-based system for real-time emulation and verification is implemented. And, from the real-time emulation results, it can be seen that the proposed algorithm has quite good performance in terms of obtaining good visual quality images.

The intention of zooming in a video camera is to obtain more detailed visual information than the full frame view. The optical zooming mechanisms provide this additional information, however, when optically zoomed, the camera no longer detects the full frame view, which was detected before zooming. Digital zooming can detect the full view at all times, however, fails to provide additional visual information when zoomed. In applications such as visual surveillance and manufacturing quality measurement, it is required to detect the full frame view at all times, the camera should be capable of providing additional visual information when zoomed to the region of interest (ROI). This paper presents a video camera system incorporating an autonomous pan, tilt and zoom functionality based on ROI decided upon motion detection capability, which provides more visual information in zoomed modes while detecting the full view at all times. The intelligent ROI decision can be made not only based on motion but also shape, color, sizes etc. of targeted objects for tracking. Image sharpening and noise reduction methods are also provided within the camera system. Standard testing results indicate that this camera system can resolve up to 800 TV lines and produces comparable color quality to other CCD based CCTV surveillance cameras, especially at good lighting conditions.

We propose a real-time stereoscopic image converter that can generate stereoscopic images with different perspective depth using motion parallax from a 2-D image. Computing the motion parallax between two adjacent 2-D images using the proposed method for motion detection, region segmentation and depth map generation generates the stereoscopic image. We have compared the stereoscopic depth of the proposed method with that of the conventional method. The proposed converter is designed and implemented by using a FPGA and verified by evaluating the field-sequential stereoscopic image on a TV. This converter can automatically produce a realistic 3-D effect regardless of the direction and velocity of a moving object in the 2-D image.

This paper presents a low power motion detection algorithm and its system architecture. The proposed architecture uses just one bit data of each selected pixel for the motion detection, where less than a few percentages of the total pixels will be selected. Since only small amount of image data is used for the motion detection, extreme low power consumption is possible. Therefore, this architecture is useful for portable battery-operated security cameras.

A motion adaptive deinterlacing algorithm is presented in this paper. It consists of the ELA-median directional interpolation, same-parity 4-field horizontal motion detection, morphological operation for noise reduction and adaptive threshold adjusting. The edges can be sharper when the ELA-median directional interpolation is adopted. The same-parity 4-field horizontal motion detection detects faster motion and makes more accurate determinations about where objects are going to move. The morphological operation for noise reduction and adaptive threshold adjusting preserve the actual texture of the original objects. The proposed method achieves cost-efficient hardware implementation with low complexity, low memory usage, and high-speed processing capability. In addition, it consumes less time in producing high-quality images and allows the audience to enjoy a high-quality TV sequence on their progressive devices. The experimental results show that the proposed algorithm is more cost-effective than previous systems.

A novel motion adaptive de-interlacing algorithm is described. The accuracy of motion detection between fields is significantly improved by difference calculation with the brightness profile pattern instead of the simple pixel differences. The performance of a pure spatial interpolation is also enhanced by directional filtering combined with a hierarchical rule based detection scheme. The proposed algorithm is successfully implemented in a (H)DTV video display processor.

In this paper, we present a new digital image stabilization (DIS) scheme based on bit-plane matching (BPM). The proposed DIS system performs motion estimation using 1-bit planes which are extracted from a video sequence. This motion estimation technique can be realized using only Boolean functions which have significantly reduced computational complexity, while the accuracy of motion estimation is maintained. In the second part of this paper, a median-based motion correction scheme is proposed which is robust to various irregular conditions such as moving objects and intentional panning. Simulation results show that the proposed DIS algorithm exhibits better performance compared with existing other algorithms when applied to real video signals.

On-line digitalization of video camera images and the processing of the entire frame within 1/25 th of the second enables complex calculation of the security functions to be performed in real time. It is therefore possible to detect the presence of new moving objects with the use of a fuzzy pattern comparator. The authors outline the time filter required to perform motion detection. The processing of the above functions is divided into two parts: those processed by the host processor and the others performed by the fuzzy pattern comparator. As there are some possibilities for parallel processing the authors show the organization of the programming system with emphasis on the sharing of tasks between the two resources. The paper shows that real-time monitoring functions are possible with general purpose and additional multimedia hardware. Performances of the system are shown through examples. >

This paper describes an advanced algorithm for motion detection from an interlaced video signal, such as PAL or NTSC. A general comparative analysis of motion detectors utilizing one, two, and three field memories is presented. As a potential application, television scan rate conversion, e.g. deinterlacing, is considered. >

In this paper a new edge and motion adaptive normal definition to higher definition (NDHD) scanning rate conversion algorithm is presented. The algorithm utilizes both edge and motion information. Edge analysis is done separately inter- and intrafield, intended for interpolation in stationary and moving regions, respectively. Motion information is obtained from a frame difference signal which is expanded horizontally if horizontal interfield edge is detected. Steerable FIR-filters are used in the interpolation. >

A deinterlacer suitable for IQTV receivers and multimedia applications is presented. The processing is based on a hierarchical motion adaptive filtering. The proposed method effectively copes with the motion present in the scene: it performs a local interpolation using data lying along an estimated direction of motion. The algorithm is synthesized using VHDL (VHSIC Hardware Description Language) into FPGA (field-programmable gate array) components and in a single programmable ASIC (application-specific integrated circuit). >

The authors propose a motion dependent VQ(MDVQ) which reduces the searching time of the VQ encoder by means of motion estimation. In a video sequence, successive frames are usually very highly correlated. Using the successive frame difference, they can classify a motion part and a still part in a frame. In MDVQ, they make two sets of codebooks. One is the main codebook which has all codewords and the other is the sub-codebook which is a subset of the main codebook. For the still part, MDVQ refers to the sub-codebook instead of the main codebook. For the motion part, it refers to the main codebook. According to computer simulation, MDVQ reduces the searching time greatly compared to normal full search VQ without any significant degradation. >

A cost-effective motion-adaptive video processing system was developed for luma/chroma separation and frame recursive noise reduction, using only a 2-Mbit frame memory. The problem of aliasing due to subsampling and the problem of waveform deterioration due to band restriction were solved, and a method of chroma motion detection using interframe difference signals was developed. Therefore, the required frame memory capacity can be reduced without a performance compromise. The video processing part of this system was produced as a single-chip digital LSI and was installed in a home-use S-VHS format VCR. The result was improved picture quality for VCR recording and playback. >

The authors describe an improved motion-adaptive de-interlacing method that incorporates the missing lines of interlaced fields from pixels of the same field or adjacent fields after motion compensation. Some improvements were obtained by suppressing the effect of false motion caused by interpolation error in the searched field, and by interpolation filtering based on the characteristics of moving objects. Experimental data show that the proposed method resulted in smaller interpolation error than conventional de-interlacing methods. >

The author describes a new algorithm for motion detection of a television signal. As possible applications, television scan rate conversions and noise reduction are considered. The proposed algorithm was not sensitive to channel noise. A memory efficient implementation is presented. >

The sampling pattern used in the HD-MAC signal depends on the motion content within the square block of pixels. There is usually a belt of blocks coded with lower resolution ahead of the moving object and on the opposite side. The authors develop a method to improve the picture quality of the erroneously coded nonmoving blocks. The additional samples needed for the high resolution are taken from the previous picture or from the following picture. >

A novel approach to digital image stabilization (DIS) for video cameras is proposed. The proposed DIS system is composed of an edge detection unit, a motion detection unit, and a digital zooming unit. The edge detection unit is based on tristate adaptive linear neurons, the motion detection unit is based on the corresponding binary logical correlation computations instead of multiple-bit pixel-wise subtractions, and the digital zooming unit is based on the approximated bilinear interpolation technique. A motion decision procedure is also proposed for removing unwanted effects on the integrated motion vector. The proposed DIS system is designed mainly for hardware minimization without using additional Micoms in a video camera system. Experimental results show that the proposed system accurately detects and compensates for camera motion. >

An algorithm for motion adaptive field rate upconversion is described. It applies one of two low-pass vertical filters to the source signal, depending on whether the source is moving or static. The theory justifying this approach is also presented. In the case of high-definition multiplexed analog component (HD-MAC) signals, the motion adaptive control can be derived from the transmitted digitally assisted TV data, while for existing standards a motion detector is required. An algorithm for motion detection is also described. To improve the quality of moving images, an additional median filtering algorithm that can be applied after the vertical filtering is presented. The display standard of 900 lines, 100 Hz/2:1 is presented as an engineering compromise, allowing the construction of 100-Hz displays using conventional techniques while avoiding any compromise of the displayed image quality. A motion adaptive field rate upconverter has been constructed that converts either 1250- or 625-line 50-Hz inputs for display at 900 lines, 100 Hz. Initial tests have shown that the concept of a high-quality receiver, as described in the text, is valid in terms of image quality. >

An electronic iron controller designed for gate array implementation is presented. For detecting the horizontal and vertical positions of the electronic iron, a motion detection input is provided for the controller. The electronic iron controller has several timing requirements for the horizontal and vertical positions. A frequency selection input pin is provided to cater to a mains frequency of 50 or 60 Hz. A buzzer output is provided by the controller for time-out purposes. >

The authors present algorithms for increasing the frame rate of an interlaced television system by a factor of 1.5. The algorithms utilize a weighted median filter structure and a simple motion detector. By adjusting the weights among the incoming fields the algorithms can be used with noninteger conversion factors. The motion detector guarantees that the performance with still pictures is optimal. The weighted median filter smoothly follows horizontal, vertical, and diagonal movement. A video sequencer was used to compare different interpolation algorithms. The proposed algorithms left few visible artifacts in the picture, and the improvement over algorithms without motion detection was particularly pronounced. The computational complexity and memory requirements of the algorithms are low. >

Circuit applications combining advanced solid- state design and manufacturing techniques have produced significant new devices for the consumer marketplace. Linear-digital devices using 12L technology, for example, led to major redesigns of conventional television receivers.

The authors have developed a digital signal processing approach for an improved NTSC (National Television System Committee) television (IDTV) system, utilizing three-dimensional motion adaptive processing techniques with frame memories. They made an effort to improve the precision of motion detection and introduced a stable system clock reproduction method and appropriate signal processing for nonstandard signals. The system has been implemented as digital LSI circuits, and an IDTV was made for home use. These digital signal processing techniques can help to improve the picture quality not only on large, direct viewing screens but also on projectors of more than 40 in. >