Interframe Transformation Fidelity Research Articles

Feature based video stabilization based on boosted HAAR Cascade and representative point matching algorithm

The success of handheld video capturing devices has further fueled the need of improved video stabilization. The videos often contain many foreground facial features like eyes, nose etc. These foreground features can be considered as feature points and may be used to stabilize videos. This paper proposes an innovative and effective digital video stabilization technique, which utilizes foreground features present in the video to produce consistent and stabilized output. It uses successive stages of Boosted HAAR cascade and representative point matching digital motion stabilization algorithm to identify and stabilize the video. The feature based tracking of object improves motion estimation accuracy between two frames thereby increasing the correlation calculation and compensation motion vector. This work achieves a significantly smoother sequence after the motion compensation. It also improves the robustness, precision and quality of the video when compared to traditional digital stabilization algorithms. The simulation results compared with pre-existing techniques reflect distinct improvements in Inter-Frame Transformation Fidelity values and Structural Similarity Index along with lesser standard deviation between image frames.

Intelligent software defined network based digital video stabilization system using frame transparency threshold pattern stabilization method

Digital Video Stabilization (DVS) allows the acquisition of video sequences without disturbing jerkiness and removing unwanted camera motion. Many researchers attempt to improve the performance of Digital Video Stabilization but all of them not get perfect results. Therefore in this research work we introduce different DVS strategy that proposes to transfer camera veins to a Frame Transparency Threshold Pattern Stabilization (FTTPS) algorithm. The proposed FTTPS algorithm consists of four steps: (i) Frame Representation; (ii) Motion Estimation; (iii) Filtering and (iv) Compensation. The proposed FTTPS method also utilizes an adaptive criterion for filtering and validation of motion vectors moreover to register a Global Motion Vector (GMV) for each frame, the proposed DVS method satisfactorily assesses the think camera motion by moving an unwanted motion of characteristics. The proposed Frame Transparency Threshold Pattern Stabilization method is basically customized in Matlab programming language, the simulation of the proposed system validated through Inter-Frame Transformation Fidelity (ITF) and Peak to Signal Noise Ratio (PSNR) Values and which is more pragmatic and takes less chip asset and general 93.12% efficiency accomplished utilizing proposed Frame Transparency Threshold Pattern Stabilization system. The suggested digital video stabilization system can be combined with sensible and robot vision, and visual monitoring structures for various low cameras display quality updates.

Bioinspired Image Stabilization Control Using the Adaptive Gain Adjustment Scheme of Vestibulo-Ocular Reflex

In this paper, an image stabilization control inspired by biological gaze stabilization mechanism is addressed. An image taken from the camera mounted on a mobile robot undergoes undesirable fluctuation that leads to poor system performance in autonomous or manual operation. Therefore, image stabilization of such camera system is essential and has been received much attention in the machine vision literature. On the other hand, a human is capable of stabilizing visual input, and this biological image stabilization mechanism is generally known as vestibulo-ocular reflex (VOR) in physiology. A particular interest of this paper is to realize this VOR features to image stabilization control problem. The mathematical interpretation of VOR in terms of the stabilization control is derived and a VOR adaptation rule is newly proposed. In the proposed method, the reference to control system is adaptively modulated by an adaptive gain to minimize residual vibration of the camera as the VOR gain converges to optimal state. To verify the proposed method, the linear shaft actuator-based image stabilization device is briefly introduced first, and it is utilized for the simulation model and experiments. Simulink simulation and experiments were conducted to demonstrate the proposed method, and results showed the proposed scheme contributes to image stabilizing effect. The image stabilization performance were validated from the vibration measurements and interframe transformation fidelity of the image sequence. From simulation and experimental results, it will be shown that the proposed method has a high potential to be applied in image stabilization control under significant system variation and uncertainties.

A compact hardware architecture for digital image stabilization using integral projections

We present a hardware architecture for real-time digital video stabilization for high-performance embedded systems. The stabilization algorithm analyzes the current and past video frames and obtains a motion estimation vector, which is then filtered to isolate unwanted camera movements from intentional panning. The vector is then used to correct the output video frame. The paper describes our hardware architecture for motion estimation, filtering and correction and its implementation on a Xilinx Spartan-6 LX45 FPGA. We evaluate our results on several benchmark video sequences, both visually and using the Inter-frame Transformation Fidelity index (ITF). Running on the 640×480-pixel video output of an infrared camera, our FPGA implementation successfully compensates involuntary camera motion at a maximum throughput of 104.15 frames per second with a 100MHz clock. The power consumption added to the FPGA by the image stabilization core is only 24.16mW.

A robust video stabilization technique using integral frame projection warping

In this paper, a video stabilization technique based on optimized dynamic time warping (DTW) has been proposed. The method uses integral frame projection warping (IFPW) for the relative motion estimation between consecutive frames. Implementation of DTW is done using dynamic programming, to give a significant reduction in terms of memory and processing time. Most of the motion estimation techniques are adversely affected due to blurring, low intensity and large displacements. The proposed method gives accurate results under these conditions and proves to be a robust for various applications such as moving platform and night shooting, over existing intensity-based motion estimation techniques. Efficiency of the proposed IFPW technique, for dark video and blurry frames, is measured using motion estimation error analysis. Overall performance evaluation of the stabilization system is done using interframe transformation fidelity and processing time.

Video stabilization using maximally stable extremal region features

Video stabilization is an important technique in present day digital cameras as most of the cameras are hand-held, mounted on moving platforms or subjected to atmospheric vibrations. In this paper we propose a novel video stabilization scheme based on estimating the camera motion using maximally stable extremal region features. These features traditionally used in wide baseline stereo problems were never explored for video stabilization purposes. Through our extensive experiments show we how some properties of these region features are suitable for the stabilization task. After estimating the global camera motion parameters using these region features, we smooth the motion parameters using a gaussian filter to retain the desired motion. Finally, motion compensation is carried out to obtain a stabilized video sequence. A number of examples on real and synthetic videos demonstrate the effectiveness of our proposed approach. We compare our results to existing techniques and show how our proposed approach compares favorably to them. Interframe Transformation Fidelity is used for objective evaluation of our proposed approach.

一种基于检测块动态选择的稳像算法

A moving object dedection-based image sabilization algorithm was propoed for addressesing the moving object problem of image sabilization.In the algorithm,the estimate global motion paramaters were used to compensate the inter-frame global motion,and the difference image between the motion compensated previous and current frames was obtained.Then,the difference image was divided into a series of forground and background blocks.And the global motion paramaters in the next run were estimated only by tracking the background blocks.Simulation results show that the proposedalgorithm has low computatiomal complexity,and achives quite high interframe transformation fidelity,which is suitable for stabilizing video clips with large moving objects.