Content based video retrival systems requires video to be segmented into objects. A large number of video segmentation algorithms have been proposed such as semi-automatic and automatic. Semiautomatic methods requires...
详细信息
Field programmable gate arrays (FPGAs) are fundamentally different to fixed processors architectures because their memory hierarchies can be tailored to the needs of an algorithm. FPGA compilers for high level languag...
详细信息
Camera sensors can only capture a limited range of luminance simultaneously, and in order to create high dynamic range (HDR) images a set of different exposures are typically combined. In this paper we address the pro...
详细信息
Camera sensors can only capture a limited range of luminance simultaneously, and in order to create high dynamic range (HDR) images a set of different exposures are typically combined. In this paper we address the problem of predicting information that have been lost in saturated image areas, in order to enable HDR reconstruction from a single exposure. We show that this problem is well-suited for deep learning algorithms, and propose a deep convolutional neural network (CNN) that is specifically designed taking into account the challenges in predicting HDR values. To train the CNN we gather a large dataset of HDR images, which we augment by simulating sensor saturation for a range of cameras. To further boost robustness, we pre-train the CNN on a simulated HDR dataset created from a subset of the MIT Places database. We demonstrate that our approach can reconstruct high-resolution visually convincing HDR results in a wide range of situations, and that it generalizes well to reconstruction of images captured with arbitrary and low-end cameras that use unknown camera response functions and post-processing. Furthermore, we compare to existing methods for HDR expansion, and show high quality results also for image based lighting. Finally, we evaluate the results in a subjective experiment performed on an HDR display. This shows that the reconstructed HDR images are visually convincing, with large improvements as compared to existing methods.
Discrete Fourier transform (DFT) finds various applications in signal processing, imageprocessing, artificial intelligent, and fuzzy logic etc. DFT is often computed efficiently with Fast Fourier transform (FF1). The...
详细信息
Discrete Fourier transform (DFT) finds various applications in signal processing, imageprocessing, artificial intelligent, and fuzzy logic etc. DFT is often computed efficiently with Fast Fourier transform (FF1). The modified split radix FFT (MSRFET.) algorithm implements a length-N=2(m) DFT achieving a reduction of arithmetic complexity compared to split-radix FF1 (SRFFT). In this paper, a simplified algorithm is proposed for the MSRFFT algorithm, reducing the number of real coefficients evaluated from 5/8N - 2 to 15/32N - 2 and the number of groups of decomposition from 4 to 3. A implementation approach is also presented. The approach makes data-path of the MSRFFT regular similar to that of the radix-2 FFT algorithm. The experimental results show that (1) MSRFFT consumes less time on central processing units (CPUs) with sufficient cache than existing algorithms;(2) the proposed implementation method can save execution time on CPUs and general processing units (GPUs).
Facial expression has made significant progress in recent years with many commercial systems are available for real-world applications. It gains strong interest to implement a facial expression system on a portable de...
详细信息
ISBN:
(纸本)9781509040933
Facial expression has made significant progress in recent years with many commercial systems are available for real-world applications. It gains strong interest to implement a facial expression system on a portable device such as tablet and smart phone device using the camera already integrated in the devices. It is very common to see face recognition phone unlocking app in new smart phones which are proven to be hassle free way to unlock a phone. Implementation a facial expression system in a smart phone would provide fun applications that can be used to measure the mood of the user in their daily life or as a tool for their daily monitoring of the motion in phycology studies. However, traditional facial expression algorithms are normally computing extensive and can only be implemented offline at a computer. In this paper, a novel automatic system has been proposed to recognize emotions from face images on a smart phone in real-time. In our system, the camera of the smart phone is used to capture the face image, BRIEF features are extracted and k-nearest neighbor algorithm is implemented for the classification. The experimental results demonstrate that the proposed facial expression recognition on mobile phone is successful and it gives up to 89.5% recognition accuracy.
In this paper, we present a novel algorithm for reliable and fast feature matching. Inspired by recent efforts in optimizing the matching process using geometric and statistical properties, we developed an approach wh...
详细信息
ISBN:
(纸本)9783319464787;9783319464770
In this paper, we present a novel algorithm for reliable and fast feature matching. Inspired by recent efforts in optimizing the matching process using geometric and statistical properties, we developed an approach which constrains the search space by utilizing spatial statistics from a small subset of matched and filtered correspondences. We call this method Guided Matching based on Statistical Optical Flow (GMbSOF). To ensure broad applicability, our approach works on high dimensional descriptors like SIFT but also on binary descriptors like FREAK. To evaluate our algorithm, we developed a novel method for determining ground truth matches, including true negatives, using spatial ground truth information of well known datasets. Therefore, we evaluate not only with precision and recall but also with accuracy and fall-out. We compare our approach in detail to several relevant state-of-the-art algorithms using these metrics. Our experiments show that our method outperforms all other tested solutions in terms of processing time while retaining a comparable level of matching quality.
This paper proposes an efficient method to improve image quality based on Context-based enhancement techniques, particularly towards real-time applications in dedicated hardware systems. The main idea is that all the ...
详细信息
This document describes a demonstration, proposed at DASIP Demo Night session. This demonstration is running on a low-power manycore processor from Kalray, named MPPA ® and is highlighting capabilities of this n...
详细信息
This document describes a demonstration, proposed at DASIP Demo Night session. This demonstration is running on a low-power manycore processor from Kalray, named MPPA ® and is highlighting capabilities of this new generation chip in processing multiple applications.
Unauthorized and unregistered sea going fishing vessels are being used for criminal activities in the coastal areas. The issue of piracy against merchant ships using illegal fishing vessels poses a significant threat ...
详细信息
ISBN:
(纸本)9781467393379
Unauthorized and unregistered sea going fishing vessels are being used for criminal activities in the coastal areas. The issue of piracy against merchant ships using illegal fishing vessels poses a significant threat to world shipping. Unfortunately counter piracy efforts and maritime security of our own and other nation enforcement often affects the innocent fishermen who conduct the trans-border fishing. Hence a proper vessel monitoring system is required to protect the maritime security without tampering the routine fishing activity of the sea going fishermen. This paper discusses about the feasibility of a system for the detection of registered marine fishing vessels comparing the satellite images and GPRS signal information. A review on various algorithms for identifying marine boats from satellite images is also conducted in this paper.
This paper describes the optical setup and imageprocessing required to estimate melt-pool width and build height for real-time control of melt-pool geometry in directed energy deposition additive manufacturing. To ov...
详细信息
暂无评论