Image processing chip for small object detection

A new, front-end image processing chip is presented for real-time small object detection. It has been implemented using a 0.6 µ, 3.3 V CMOS technology and operates on 10-bit input data at 54 megasamples per second. It occupies an area of 12.9 mm×13.6 mm (including pads), dissipates 1.5 W, has 92 I/O pins and is to be housed in a 160-pin ceramic quarter flat-pack. It performs both one- and two-dimensional FIR filtering and a multilayer perceptron (MLP) neural network function using a reconfigurable array of 21 multiplication-accumulation cells which corresponds to a window size of 7×3. The chip can cope with images of 2047 pixels per line and can be cascaded to cope with larger window sizes. The chip performs two billion fixed point multiplications and additions per second.

Interferometer for Co-operating Target Motion Detection

A correlation interferometer working with a cooperation target operating at 2.2GHz center frequency is presented. This simplified interferometer presented here uses a
lock-in amplifier to significantly increase system sensitivity when used in conjunction with a co-operating target signaling using amplitude modulation. The system is verified by detecting the angular velocity of passing tagged target. Experimental results show detectable range up over 110 meters in a multipath environment using a 10dBm EIRP tag.

Semantic feature-based visual attention model for pedestrian detection

Objective
Pedestrian detection under video surveillance systems has always been a hot topic in computer vision research. These systems are widely used in train stations, airports, large commercial plazas, and other public places. However, pedestrian detection remains difficult because of complex backgrounds. Given its development in recent years, the visual attention mechanism has attracted increasing attention in object detection and tracking research, and previous studies have achieved substantial progress and breakthroughs. We propose a novel pedestrian detection method based on the semantic features under the visual attention mechanism.
Method
The proposed semantic feature-based visual attention model is a spatial-temporal model that consists of two parts: the static visual attention model and the motion visual attention model. The static visual attention model in the spatial domain is constructed by combining bottom-up with top-down attention guidance. Based on the characteristics of pedestrians, the bottom-up visual attention model of Itti is improved by intensifying the orientation vectors of elementary visual features to make the visual saliency map suitable for pedestrian detection. In terms of pedestrian attributes, skin color is selected as a semantic feature for pedestrian detection. The regional and Gaussian models are adopted to construct the skin color model. Skin feature-based visual attention guidance is then proposed to complete the top-down process. The bottom-up and top-down visual attentions are linearly combined using the proper weights obtained from experiments to construct the static visual attention model in the spatial domain. The spatial-temporal visual attention model is then constructed via the motion features in the temporal domain. Based on the static visual attention model in the spatial domain, the frame difference method is combined with optical flowing to detect motion vectors. Filtering is applied to process the field of motion vectors. The saliency of motion vectors can be evaluated via motion entropy to make the selected motion feature more suitable for the spatial-temporal visual attention model.
Result
Standard datasets and practical videos are selected for the experiments. The experiments are performed on a MATLAB R2012a platform. The experimental results show that our spatial-temporal visual attention model demonstrates favorable robustness under various scenes, including indoor train station surveillance videos and outdoor scenes with swaying leaves. Our proposed model outperforms the visual attention model of Itti, the graph-based visual saliency model, the phase spectrum of quaternion Fourier transform model, and the motion channel model of Liu in terms of pedestrian detection. The proposed model achieves a 93% accuracy rate on the test video.
Conclusion
This paper proposes a novel pedestrian method based on the visual attention mechanism. A spatial-temporal visual attention model that uses low-level and semantic features is proposed to calculate the saliency map. Based on this model, the pedestrian targets can be detected through focus of attention shifts. The experimental results verify the effectiveness of the proposed attention model for detecting pedestrians.

Programmable image processing chip

A new high performance, programmable image processing chip targeted at video and HDTV applications is described. This was initially developed for image small object recognition but has much broader functional application including 1D and 2D FIR filtering as well as neural network computation. The core of the circuit is made up of an array of twenty one multiplication-accumulation cells based on systolic architecture. Devices can be cascaded to increase the order of the filter both vertically and horizontally. The chip has been fabricated in a 0.6 µ, low power CMOS technology and operates on 10 bit input data at over 54 Megasamples per second. The introduction gives some background to the chip design and highlights that there are few other comparable devices. Section 2 gives a brief introduction to small object detection. The chip architecture and the chip design will be described in detail in the later sections.

Histogram of oriented gradients front end processing: An FPGA based processor approach

The Field Programmable Gate Array (FPGA) implementation of the commonly used Histogram of Oriented Gradients (HOG) algorithm is explored. The HOG algorithm is employed to extract features for object detection. A key focus has been to explore the use of a new FPGA-based processor which has been targeted at image processing. The paper gives details of the mapping and scheduling factors that influence the performance and the stages that were undertaken to allow the algorithm to be deployed on FPGA hardware, whilst taking into account the specific IPPro architecture features. We show that multi-core IPPro performance can exceed that of against state-of-the-art FPGA designs by up to 3.2 times with reduced design and implementation effort and increased flexibility all on a low cost, Zynq programmable system.

Automated detection and tracking of solar magnetic bright points

Magnetic bright points (MBPs) in the internetwork are among the smallest objects in the solar photosphere and appear bright against the ambient environment. An algorithm is presented that can be used for the automated detection of the MBPs in the spatial and temporal domains. The algorithm works by mapping the lanes through intensity thresholding. A compass search, combined with a study of the intensity gradient across the detected objects, allows the disentanglement of MBPs from bright pixels within the granules. Object growing is implemented to account for any pixels that might have been removed when mapping the lanes. The images are stabilized by locating long-lived objects that may have been missed due to variable light levels and seeing quality. Tests of the algorithm, employing data taken with the Swedish Solar Telescope, reveal that approximate to 90 per cent of MBPs within a 75 x 75 arcsec(2) field of view are detected.

Gliadin antibody detection in gluten enteropathy

Circulating antigliadin antibody has been described in patients with gluten enteropathy although the prevalence varies in different studies. It has been suggested that the investigation for antigliadin antibody might be useful as a screening test. The object of the present study was to evaluate two different techniques for assaying these antibodies - an indirect immunofluorescent method and an enzyme-linked immunosorbent assay (ELISA). Antibodies were assayed in the sera of 102 patients in whom jejunal biopsies were also obtained. The specificity of both tests was greater than 95%, and the correlation between the presence of antibody and histology was significant (p <0.005), though the sensitivity of each test was less than 70%.

The Pan-STARRS Moving Object Processing System

We describe the Pan-STARRS Moving Object Processing System (MOPS), a modern software package that produces automatic asteroid discoveries and identifications from catalogs of transient detections from next-generation astronomical survey telescopes. MOPS achieves >99.5% efficiency in producing orbits from a synthetic but realistic population of asteroids whose measurements were simulated for a Pan-STARRS4-class telescope. Additionally, using a nonphysical grid population, we demonstrate that MOPS can detect populations of currently unknown objects such as interstellar asteroids. MOPS has been adapted successfully to the prototype Pan-STARRS1 telescope despite differences in expected false detection rates, fill-factor loss, and relatively sparse observing cadence compared to a hypothetical Pan-STARRS4 telescope and survey. MOPS remains highly efficient at detecting objects but drops to 80% efficiency at producing orbits. This loss is primarily due to configurable MOPS processing limits that are not yet tuned for the Pan-STARRS1 mission. The core MOPS software package is the product of more than 15 person-years of software development and incorporates countless additional years of effort in third-party software to perform lower-level functions such as spatial searching or orbit determination. We describe the high-level design of MOPS and essential subcomponents, the suitability of MOPS for other survey programs, and suggest a road map for future MOPS development.

Video anomaly detection in real-time on a Power-Aware Heterogeneous Platform

FPGAs and GPUs are often used when real-time performance in video processing is required. An accelerated processor is chosen based on task-specific priorities (power consumption, processing time and detection accuracy), and this decision is normally made once at design time. All three characteristics are important, particularly in battery-powered systems. Here we propose a method for moving selection of processing platform from a single design-time choice to a continuous run time one.We implement Histogram of Oriented Gradients (HOG) detectors for cars and people and Mixture of Gaussians (MoG) motion detectors running across FPGA, GPU and CPU in a heterogeneous system. We use this to detect illegally parked vehicles in urban scenes. Power, time and accuracy information for each detector is characterised. An anomaly measure is assigned to each detected object based on its trajectory and location, when compared to learned contextual movement patterns. This drives processor and implementation selection, so that scenes with high behavioural anomalies are processed with faster but more power hungry implementations, but routine or static time periods are processed with power-optimised, less accurate, slower versions. Real-time performance is evaluated on video datasets including i-LIDS. Compared to power-optimised static selection, automatic dynamic implementation mapping is 10% more accurate but draws 12W extra power in our testbed desktop system.