ASP-SIFT: Using Analog Signal Processing Architecture to Accelerate Keypoint Detection of SIFT Algorithm

Scale invariant feature transform(SIFT) is an algorithm to extract distinctive and invariant features from images to achieve reliable object matching between different images in variant scales and rotations. However, SIFT's huge computation impedes its real-time implementation. In this paper, a Fast-Gaussian SIFT(FG-SIFT) is proposed. Keypoint detection is optimized in FG-SIFT. SIFT's 2-D difference of Gaussian(DoG) in Gaussian Scale-Space(GSS) is separated into two 1-D DoG in x and y dimensions, and the level of scales in DoG pyramid is also reduced. The experiment shows that FG-SIFT reduces the computational complexity about 95% in GSS construction, also increases the accuracy of keypoint detection. Subsequently, the accuracy of generated features is increased 162%, and the accuracy of matched features is increased 8%. Apart from optimization in algorithm level, a hardware architecture of FG-SIFT's keypoint detection module is proposed. With a parallel architectural incorporating a five-stage pipeline, the execution time of keypoint detection is only 1.42ms@Xilinx Virtex5. Compared to conventional works, the speed is 21% faster than the fastest solution [9](ASIC), and hardware resources are 70% less than the most resources saved solutions [5] [6](Xilinx Virtex5).

Fully pipelined FPGA-based architecture for real-time SIFT extraction

Scale-invariance, good localization and robustness to noise and distortions are the main properties that a local feature detector should possess. Most existing local feature detectors find excessive unstable feature points that increase the number of keypoints to be matched and the computational time of the matching step. In this paper, we show that robust and accurate keypoints exist in the specific scale-space domain. To this end, we first formulate the superimposition problem into a mathematical model and then derive a closed-form solution for multiscale analysis. The model is formulated via difference-of-Gaussian (DoG) kernels in the continuous scale-space domain, and it is proved that setting the scale-space pyramid's blurring ratio and smoothness to 2 and 0.627, respectively, facilitates the detection of reliable keypoints. For the applicability of the proposed model to discrete images, we discretize it using the undecimated wavelet transform and the cubic spline function. Theoretically, the complexity of our method is less than 5% of that of the popular baseline Scale Invariant Feature Transform (SIFT). Extensive experimental results show the superiority of the proposed feature detector over the existing representative hand-crafted and learning-based techniques in accuracy and computational time. The code and supplementary materials can be found at https://github.com/mogvision/FFD.

SIFT (scale invariant feature transform) is one of the most robust and widely used image matching algorithms based on local features. However, its computational complexity is high. In order to reduce the matching time, an improved feature matching algorithm is proposed in this paper under the premise of stable registration accuracy. This paper proposed a normalized cross-correlation with SIFT combination of remote sensing image matching algorithm. The basic idea of the algorithm is performing the space geometry transformation of the input image with reference to the base image. Then the normalized cross-correlation captures the relevant part of the remote sensing images. By this way, we can reduce the matching range. So some unnecessary calculations are properly omitted. By utilizing the SIFT algorithm, we match the preprocessed remote sensing images, and get the registration points. This can shorten the matching time and improve the matching accuracy. Its robustness is increased correspondingly. The experimental results show that the proposed Normalized cross-correlation plus SIFT algorithm is more rapid than the standard SIFT algorithm while the performance is favorably compared to the standard SIFT algorithm when matching among structured scene images. The experiment results confirm the feasibility of our methods.

We evaluate the performance of different key-point detectors and region descriptors when used for automatic classification of small sea targets in infrared video. In our earlier research performed on this subject as well as in other literature, many different region descriptors have been proposed. However, it is unclear which methods are most applicable to use on the type of infrared imagery as used onboard naval ships. The key-point detector should detect points of interest that can be used to effectively describe the objects in the imagery. On the basis of the detected key points, the descriptors should discriminate between different classes of small sea targets while being robust to differences in viewing conditions. We propose a similarity measure based on the distance between key-point location and the Euclidean distance between descriptors to quantify the similarity of images. For performance evaluation, we use the receiver operator characteristic as the criterion to rank the evaluated methods. We compare the Harris-, blob- and scale-invariant feature transform (SIFT) detectors and the square neighborhood, steerable filters, invariant moments, and SIFT descriptors. We conclude that the Harris detector combined with the square neighborhood of size 19×19 or the SIFT descriptor results in the best classification performance for our data set.

In recent years, there has been an increasing requirement for methods of detecting bubbles released from the seabed into the water column, such as leaks from undersea gas pipelines and seeps from carbon capture and storage facilities. Considering the peculiarity of the layout of submarine gas pipelines and that of the ocean environment around them, we construct an underwater mobile platform equipped with autonomous underwater vehicles carrying multi-beam echo sounder (MBES) and various types of other sensors. Analogous to optical flow, this paper describes a scale-invariant feature transform (SIFT) flow algorithm, which includes both the detection of key points and the computation of local descriptors. This SIFT flow algorithm estimates the motion characteristics of gas leaks and the results show a positive correlation relationship between the rising velocity of leakage gas and the leakage pressure as well as the leakage aperture. Finally, the validity of this method is verified in tank and sea experimental research.

The SIFT (scale invariant feature transform) has demonstrated its superior performance in identifying transformed images over many other approaches. However, both of its detection and matching stages are expensive, because a large number of keypoints are detected in the scale-space and each keypoint is described using a 128-dimensional vector. We present two possible solutions for feature-point reduction. First is to down scale the image before the SIFT keypoint detection and second is to use corners (instead of SIFT keypoints) which are visually significant, more robust, and much smaller in number than the SIFT keypoints. Either the curvature descriptor or the highly distinctive SIFT descriptors at corner locations can be used to represent corners.We then describe a new feature-point matching technique, which can be used for matching both the down-scaled SIFT keypoints and corners. Experimental results show that two feature-point reduction solutions combined with the SIFT descriptors and the proposed feature-point matching technique not only improve the computational efficiency and decrease the storage requirement, but also improve the transformed image identification accuracy (robustness).

To improve the efficiency and effectiveness of mosaicking unmanned aerial vehicle (UAV) images, we propose in this paper a rapid mosaicking method based on scale-invariant feature transform (SIFT) for mosaicking UAV images used for crop growth monitoring. The proposed method dynamically sets the appropriate contrast threshold in the difference of Gaussian (DOG) scale-space according to the contrast characteristics of UAV images used for crop growth monitoring. Therefore, this method adjusts and optimizes the number of matched feature point pairs in UAV images and increases the mosaicking efficiency. Meanwhile, based on the relative location relationship of UAV images used for crop growth monitoring, the random sample consensus (RANSAC) algorithm is integrated to eliminate the influence of mismatched point pairs in UAV images on mosaicking and to keep the accuracy and quality of mosaicking. Mosaicking experiments were conducted by setting three types of UAV images in crop growth monitoring: visible, near-infrared, and thermal infrared. The results indicate that compared to the standard SIFT algorithm and frequently used commercial mosaicking software, the method proposed here significantly improves the applicability, efficiency, and accuracy of mosaicking UAV images in crop growth monitoring. In comparison with image mosaicking based on the standard SIFT algorithm, the time efficiency of the proposed method is higher by 30%, and its structural similarity index of mosaicking accuracy is about 0.9. Meanwhile, the approach successfully mosaics low-resolution UAV images used for crop growth monitoring and improves the applicability of the SIFT algorithm, providing a technical reference for UAV application used for crop growth and phenotypic monitoring.

This paper presents an automated image registration approach to detecting changes in multi-temporal remote sensing images. The proposed algorithm is based on the scale invariant feature transform (SIFT) and has two phases. The first phase focuses on SIFT feature extraction and on estimation of image transformation. In the second phase, Structured Local Binary Haar Pattern (SLBHP) combined with a fuzzy similarity measure is then used to build a new and effective block similarity measure for change detection. Experimental results obtained on multi-temporal data sets show that compared with three mainstream block matching algorithms, the proposed algorithm is more effective in dealing with scale, rotation and illumination changes.

There has been an extensive growth in the area of digital image processing. One of the main issues in the real world is to judge the authenticity of a given image. Nowadays it is very easy to tamper and forge digital image with the advancement of the powerful digital image processing software and easy availability of the tools. The most common form of image manipulation techniques is the region duplication also called as copymove forgery where a portion of the image is copied and paste to another portion in the same digital image. To investigate such forensic analysis, various techniques and method have been developed in the past literature. There has been an immense research of finding such tampered pixel blocks. This method cannot handle cases when the copied region is scaled or rotated to a new region. This paper proposes an efficient algorithm based on scale invariant feature transform (SIFT) with feature extraction which is invariant to translation, scale, noise and rotation. It comprises transformation of the input image to produce a standard representation and then detection of keypoint and feature descriptor is applied along with a matching over all the keypoints.

To improve the real-time performance, a quick scale-invariant interest point detecting approach based on the image color information is proposed in this paper. The approach uses the scale normalized Laplacian operator to extract the interest points in the incomplete image pyramid. A new local descriptor is presented in the approach to compute the feature vector of each interest point. The descriptor is made up with several subregions like the SIFT (Scale-Invariant Feature Transform) descriptor, meanwhile, it chooses the mean values of different color components in each subregion as the feature vector’s elements to differentiate color objects better and reduce the descriptor’s dimension. Through the experiment, the detected interest points are robust to many image transformations and the approach is indicative of needing less computation than other interest point detecting algorithms. The research discloses that the approach can obtain both superior stability and real-time performance at the same time.

Nowadays advantages in face-based modification using DeepFake algorithms made it possible to replace a face of one person with a face of another person. Thus, it is possible to make not only copy-move modifications, but to implement artificial intelligence and deep learning for replacing face movements from one person to another. Still images can be converted into video sequences. Consequently, the contemporaries, historical figures or even animated characters can be lively presented. Deepfakes are becoming more and more successful and it is difficult to detect them in some cases. In this paper we explain the video sequences we produced (e.g. using X2Face method, and First Order Motion Model for Image Animation) and perform deepfake video analysis using SIFT (Scale Invariant Feature Transform) based approach. The experiments show the simplicity in video forgery production, as well as the possible role of SIFT keypoints detection in differentiation between the deeply forged and original video content.

Wide-angle images exhibit significant distortion for which existing scale-space detectors such as the scale-invariant feature transform (SIFT) are inappropriate. The required scale-space images for feature detection are correctly obtained through the convolution of the image, mapped to the sphere, with the spherical Gaussian. A new visual key-point detector, based on this principle, is developed and several computational approaches to the convolution are investigated in both the spatial and frequency domain. In particular, a close approximation is developed that has comparable computation time to conventional SIFT but with improved matching performance. Results are presented for monocular wide-angle outdoor image sequences obtained using fisheye and equiangular catadioptric cameras. We evaluate the overall matching performance (recall versus 1-precision) of these methods compared to conventional SIFT. We also demonstrate the use of the technique for variable frame-rate visual odometry and its application to place recognition.

The synthesis of real filters using the complex coefficient filters is considered interesting in digital signal processing, since one can expect that the freedom in the design will be increased and the coefficient sensitivity will be decreased. In analog signal processing, on the other hand, there are few reports on the design of the filter from a similar viewpoint.This paper discusses the complex coefficient filter on the analog domain, which is obtained by shifting the frequency response of the lower‐order real filter on the frequency axis, together with its application to synthesize real filters.First, the complex analog transfer function is introduced, which is obtained by shifting the frequency response of the real filter on the frequency axis. The elements for the complex analog signal processing are defined. Then, using those elements, the complex resonator is proposed.The proposed circuit can be considered as a special class of the biquad circuit, where the complex conjugate poles of the second‐order section are specified. As a result of sensitivity analysis, the proposed circuit is shown to have the same or lower sensitivity as that of the biquad circuit. Another feature is that the circuit can be realized using the lossy integrators and can be used as the functional unit in the active filter to be operated at higher frequencies. Design examples using the operational amplifier and their experimental results are shown to verify the effectiveness of the proposed method.

In this work we will work on analogue signal processing in the neural circuit of C. elegans which is able to detect the analogue signals from the environment and produce locomotive behaviours which are in accordance with experiments. The signals in C. elegans are processed in a purely analogue procedure, since no action potential has been recorded in its neural activity. We aim to show how signal processing can be executed in analogue domain in a living creature. In order to do that we will model two different behaviours of C. elegans which are generated in the same network of neurons, klinotaxis behaviour and isothermal tracking. We will implement a Genetic Algorithm to find appropriate sets of parameters of the model. Our contribution is to show how relatively straight forward differential equations can lead to relatively complex and different behaviours.