Search

With the exponential growth of remote sensing images in recent years, there has been a significant increase in demand for micro-target detection. Recently, effective detection methods for small targets have emerged; however, for micro-targets (even fewer pixels than small targets), most existing methods are not fully competent in feature extraction, target positioning, and rapid classification. This study proposes an enhanced detection method, especially for micro-targets, in which a combined loss function (consisting of NWD and CIOU) is used instead of a singular CIOU loss function. In addition, the lightweight Content-Aware Reassembly of Features (CARAFE) replaces the original bilinear interpolation upsampling algorithm, and a spatial pyramid structure is added into the network model’s small target layer. The proposed algorithm undergoes training and validation utilizing the benchmark dataset known as AI-TOD. Compared to speed-oriented YOLOv7-tiny, the mAP0.5 and mAP0.5:0.95 of our improved algorithm increased from 42.0% and 16.8% to 48.7% and 18.9%, representing improvements of 6.7% and 2.1%, respectively, while the detection speed was almost equal to that of YOLOv7-tiny. Furthermore, our method was also tested on a dataset of multi-scale targets, which contains small targets, medium targets, and large targets. The results demonstrated that mAP0.5:0.95 increased from “9.8%, 54.8%, and 68.2%” to “12.6%, 55.6%, and 70.1%” for detection across different scales, indicating improvements of 2.8%, 0.8%, and 1.9%, respectively. In summary, the presented method improves detection metrics for micro-targets in various scenarios while satisfying the requirements of detection speed in a real-time system., Article; Export Date: 06 March 2024; Cited By: 0; Correspondence Address: P. Wu; School of Computer and Information Technology, Beijing Jiaotong University, Beijing, 100044, China; email: 14112078@bjtu.edu.cn

Infrared small target detection is widely used in fields such as military and security. UNet, which is a classical semantic segmentation method proposed in 2015, has shown excellent performance and robustness. However, U-Net suffers from the problem of losing small targets in deep layers after multiple down-sampling operations. Dilated convolution, as a special convolution that can increase the receptive field without increasing the number of parameters, is considered to be able to optimize the problems caused by down-sampling. Dense Nested Attention Network (DNANet), due to its superior performance, was chosen as the baseline, but it still has the issue of target loss. This study proposes three optimization directions: deep down-sampling replaced by cascaded dilated convolution, dilated spatial attention, and dilated residual block. In these three directions, this study proposes four methods, respectively DNANet-DS-1, DNANet-DS-2, DNANet-Att, and DNANet-RB. Two open-source infrared small target datasets, NUDT-SIRST and NUAA-SIRST, were used in this study. The four proposed methods were trained and tested on these two datasets. Among them, DNANetRB significantly outperforms other methods on the NUAA-SIRST dataset, so further experiments were conducted to observe the influence of different network depths on DNANet-RB. The experimental result indicates that when the network depth exceeds a certain threshold, the network can only achieve marginal improvements, but the number of parameters will increase significantly., Infraröd detektering av små mål används ofta inom områden som militär och säkerhet. U-Net, som är en klassisk semantisk segmenteringsmetod som föreslogs 2015, har visat utmärkt prestanda och robusthet. U-Net lider dock av problemet med att förlora små mål i djupa lager efter flera nedprovningsoperationer. Dilaterad konvolution, som är en speciell konvolution som kan öka det receptiva fältet utan att öka antalet parametrar, anses kunna optimera de problem som orsakas av downsampling. DNANet (Dense Nested Attention Network) valdes som baslinje på grund av dess överlägsna prestanda, men det har fortfarande problemet med målförlust. Denna studie föreslår tre optimeringsriktningar: djup nedsampling ersatt av kaskad dilaterad konvolution, dilaterad rumslig uppmärksamhet och dilaterat restblock. I dessa tre riktningar föreslår denna studie fyra metoder, respektive DNANet-DS-1, DNANet-DS-2, DNANet-Att och DNANet-RB. Två dataset med små infraröda mål med öppen källkod, NUDT-SIRST och NUAA-SIRST, användes i denna studie. De fyra föreslagna metoderna tränades och testades på dessa två datamängder. Bland dem överträffar DNANet-RB betydligt andra metoder på NUAA-SIRST-datasetet, så ytterligare experiment genomfördes för att observera påverkan av olika nätverksdjup på DNANet-RB. Det experimentella resultatet visar att när nätverksdjupet överskrider ett visst tröskelvärde kan nätverket bara uppnå marginella förbättringar, men antalet parametrar kommer att öka avsevärt.

Infrared small target detection is widely used in fields such as military and security. UNet, which is a classical semantic segmentation method proposed in 2015, has shown excellent performance and robustness. However, U-Net suffers from the problem of losing small targets in deep layers after multiple down-sampling operations. Dilated convolution, as a special convolution that can increase the receptive field without increasing the number of parameters, is considered to be able to optimize the problems caused by down-sampling. Dense Nested Attention Network (DNANet), due to its superior performance, was chosen as the baseline, but it still has the issue of target loss. This study proposes three optimization directions: deep down-sampling replaced by cascaded dilated convolution, dilated spatial attention, and dilated residual block. In these three directions, this study proposes four methods, respectively DNANet-DS-1, DNANet-DS-2, DNANet-Att, and DNANet-RB. Two open-source infrared small target datasets, NUDT-SIRST and NUAA-SIRST, were used in this study. The four proposed methods were trained and tested on these two datasets. Among them, DNANetRB significantly outperforms other methods on the NUAA-SIRST dataset, so further experiments were conducted to observe the influence of different network depths on DNANet-RB. The experimental result indicates that when the network depth exceeds a certain threshold, the network can only achieve marginal improvements, but the number of parameters will increase significantly., Infraröd detektering av små mål används ofta inom områden som militär och säkerhet. U-Net, som är en klassisk semantisk segmenteringsmetod som föreslogs 2015, har visat utmärkt prestanda och robusthet. U-Net lider dock av problemet med att förlora små mål i djupa lager efter flera nedprovningsoperationer. Dilaterad konvolution, som är en speciell konvolution som kan öka det receptiva fältet utan att öka antalet parametrar, anses kunna optimera de problem som orsakas av downsampling. DNANet (Dense Nested Attention Network) valdes som baslinje på grund av dess överlägsna prestanda, men det har fortfarande problemet med målförlust. Denna studie föreslår tre optimeringsriktningar: djup nedsampling ersatt av kaskad dilaterad konvolution, dilaterad rumslig uppmärksamhet och dilaterat restblock. I dessa tre riktningar föreslår denna studie fyra metoder, respektive DNANet-DS-1, DNANet-DS-2, DNANet-Att och DNANet-RB. Två dataset med små infraröda mål med öppen källkod, NUDT-SIRST och NUAA-SIRST, användes i denna studie. De fyra föreslagna metoderna tränades och testades på dessa två datamängder. Bland dem överträffar DNANet-RB betydligt andra metoder på NUAA-SIRST-datasetet, så ytterligare experiment genomfördes för att observera påverkan av olika nätverksdjup på DNANet-RB. Det experimentella resultatet visar att när nätverksdjupet överskrider ett visst tröskelvärde kan nätverket bara uppnå marginella förbättringar, men antalet parametrar kommer att öka avsevärt.

Fitt's law is commonly used to model target selection. But Fitts' law deals with only one kind of selection strategy. Our question is, do changes in target size, distance and direction to a target affect the differences in performance between target selection strategies? We performed the first empirical tests on a pen-based system to evaluate differences in performance between six selection strategies for selecting a target. Three target sizes, eight pen-movement-directions and three pen-movement-distances were applied to all six strategies. The results show that differences between selection strategies are affected by variations in target size but not by the other parameters (distance and direction).

An experiment is reported comparing six pen input strategies for selecting a small target using five diffenent sized targets (1, 3, 5, 7 and 9 dot diameter circles respectively, 0. 36 mm per dot). The results showed that the best strategy, in terms of error rate, selection time and subjective preferences, was the "land-on2" strategy where the target is selected when the pen-tip touches the target for the first time after landing on the screen surface. Moreover, "the smallest maximum size" was determined to be 5 dots (1. 8 mm). This was the largest size among the targets which had a significant main effect on error rate in the six strategies. These results are important for both researchers and designers of pen-based systems.

We present an algorithm that produces a fused false color representation of a combined multiband IR and visual imaging system for maritime applications. Multispectral IR imaging techniques are increasingly deployed in maritime operations, to detect floating mines or to find small dinghies and swimmers during search and rescue operations. However, maritime backgrounds usually contain a large amount of clutter that severely hampers the detection of small targets. Our new algorithm deploys the correlation between the target signatures in two different IR frequency bands (3-5 and 8-12 ìm) to construct a fused IR image with a reduced amount of clutter. The fused IR image is then combined with a visual image in a false color RGB representation for display to a human operator. The algorithm works as follows. First, both individual IR bands are filtered with a morphological opening top-hat transform to extract small details. Second, a common image is extracted from the two filtered IR bands, and assigned to the red channel of an RGB image. Regions of interest that appear in both IR bands remain in this common image, while most uncorrelated noise details are filtered out. Third, the visual band is assigned to the green channel and, after multiplication with a constant (typically 1.6) also to the blue channel. Fourth, the brightness and colors of this intermediate false color image are renormalized by adjusting its first order statistics to those of a representative reference scene. The result of these four steps is a fused color image, with naturalistic colors (bluish sky and grayish water), in which small targets are clearly visible. Keywords: target detection; false color fusion, small targets, clutter, image fusion

A method is provided for estimating the size of an object from a region of a fluid medium when that object is emitting acoustic radiation of-known wavelength lambda on its own or as the result of being interrogated by acoustic pulses that reflect from the object. The acoustic radiation is monitored using a line array of N acoustic receivers such that N signals indicative of the acoustic radiation are generated. M time series summations are formed using the N signals. Each of the M time series summations is formed using a unique time delay predicated on a corresponding unique estimated speed of propagation of the acoustic radiation where M estimated speeds of propagation are defined. For an object in the region having a diameter D on the order of lambda, the M values will vary as a function of the M estimated speeds of propagation with the resulting distribution of the M values being indicative of diameter D.

Fitt's law is commonly used to model target selection. But Fitts' law deals with only one kind of selection strategy. Our question is, do changes in target size, distance and direction to a target affect the differences in performance between target selection strategies? We performed the first empirical tests on a pen-based system to evaluate differences in performance between six selection strategies for selecting a target. Three target sizes, eight pen-movement-directions and three pen-movement-distances were applied to all six strategies. The results show that differences between selection strategies are affected by variations in target size but not by the other parameters (distance and direction)., journal article

An experiment is reported comparing six pen input strategies for selecting a small target using five diffenent sized targets (1, 3, 5, 7 and 9 dot diameter circles respectively, 0. 36 mm per dot). The results showed that the best strategy, in terms of error rate, selection time and subjective preferences, was the "land-on2" strategy where the target is selected when the pen-tip touches the target for the first time after landing on the screen surface. Moreover, "the smallest maximum size" was determined to be 5 dots (1. 8 mm). This was the largest size among the targets which had a significant main effect on error rate in the six strategies. These results are important for both researchers and designers of pen-based systems., journal article