The Effect of Polarity on Object Recognition in Thermal Images

Author(s):  
Michael S. Brickner ◽  
Amir Zvuloni

Thermal imaging (TI) systems, transform the distribution of relative temperatures in a scene into a visible TV image. TIs differ significantly from regular TV images. Most TI systems allow their operators to select preferred polarity which determines the way in which gray shades represent different temperatures. Polarity may be set to either black hot (BH) or white hot (WH). The present experiments were designed to investigate the effects of polarity on object recognition performance in TI and to compare object recognition performance of experts and novices. In the first experiment, twenty flight candidates were asked to recognize target objects in 60 dynamic TI recordings taken from two different TI systems. The targets included a variety of human placed and natural objects. Each subject viewed half the targets in BH and the other half in WH polarity in a balanced experimental design. For 24 out of the 60 targets one direction of polarity produced better performance than the other. Although the direction of superior polarity (BH or WH better) was not consistent, the preferred representation of the target object was very consistent. For example, vegetation was more readily recognized when presented as dark objects on a brighter background. The results are discussed in terms of importance of surface determinants versus edge determinants in the recognition of TI objects. In the second experiment, the performance of 10 expert TI users was found to be significantly more accurate but not much faster than the performance of 20 novice subjects.

Author(s):  
Sathish K. Gurupatham ◽  
Erhan Ilksoy ◽  
Nick Jacob ◽  
Kevin Van Der Horn ◽  
Fahad Fahad

Novel technologies have always been an indispensable part of the scientific enterprise and a catalyst for new discoveries. The invisible radiation patterns of objects are converted into visible images called thermograms or thermal images. Thermal images can be utilized to estimate the ripeness of some fruits which do not change their color from yellow to green when they are ripe. Thermal imaging techniques are very helpful since color and fluorescent analytical approaches cannot be applied to these fruits. In this work, it is shown that different ripeness levels of avocado (Hall type) using a non-destructive method called thermal imaging, in two dimensional spaces. The work is based on the fact that fruits have different specific heat capacities at different temperatures, thus making their thermal images clear indicators of ripeness.


2021 ◽  
Vol 7 (4) ◽  
pp. 65
Author(s):  
Daniel Silva ◽  
Armando Sousa ◽  
Valter Costa

Object recognition represents the ability of a system to identify objects, humans or animals in images. Within this domain, this work presents a comparative analysis among different classification methods aiming at Tactode tile recognition. The covered methods include: (i) machine learning with HOG and SVM; (ii) deep learning with CNNs such as VGG16, VGG19, ResNet152, MobileNetV2, SSD and YOLOv4; (iii) matching of handcrafted features with SIFT, SURF, BRISK and ORB; and (iv) template matching. A dataset was created to train learning-based methods (i and ii), and with respect to the other methods (iii and iv), a template dataset was used. To evaluate the performance of the recognition methods, two test datasets were built: tactode_small and tactode_big, which consisted of 288 and 12,000 images, holding 2784 and 96,000 regions of interest for classification, respectively. SSD and YOLOv4 were the worst methods for their domain, whereas ResNet152 and MobileNetV2 showed that they were strong recognition methods. SURF, ORB and BRISK demonstrated great recognition performance, while SIFT was the worst of this type of method. The methods based on template matching attained reasonable recognition results, falling behind most other methods. The top three methods of this study were: VGG16 with an accuracy of 99.96% and 99.95% for tactode_small and tactode_big, respectively; VGG19 with an accuracy of 99.96% and 99.68% for the same datasets; and HOG and SVM, which reached an accuracy of 99.93% for tactode_small and 99.86% for tactode_big, while at the same time presenting average execution times of 0.323 s and 0.232 s on the respective datasets, being the fastest method overall. This work demonstrated that VGG16 was the best choice for this case study, since it minimised the misclassifications for both test datasets.


Perception ◽  
1998 ◽  
Vol 27 (1) ◽  
pp. 47-68 ◽  
Author(s):  
Fiona N Newell

The effect of stimulus factors such as interobject similarity and stimulus density on the recognition of objects across changes in view was investigated in five experiments. The recognition of objects across views was found to depend on the degree of interobject similarity and on stimulus density: recognition was view dependent when both interobject similarity and stimulus density were high, irrespective of the familiarity of the target object. However, when stimulus density or interobject similarity was low recognition was invariant to viewpoint. It was found that recognition was accomplished through view-dependent procedures when discriminability between objects was low. The findings are discussed in terms of an exemplar-based model in which the dimensions used for discriminating between objects are optimised to maximise the differences between the objects. This optimisation process is characterised as a perceptual ‘ruler’ which measures interobject similarity by stretching across objects in representational space. It is proposed that the ‘ruler’ optimises the feature differences between objects in such a way that recognition is view invariant but that such a process incurs a cost in discriminating between small feature differences, which results in view-dependent recognition performance.


2021 ◽  
Vol 11 (5) ◽  
pp. 2358
Author(s):  
Mitsuhiko Kimoto ◽  
Takamasa Iio ◽  
Masahiro Shiomi ◽  
Katsunori Shimohara

This study proposes a robot conversation strategy involving speech and gestures to improve a robot’s indicated object recognition, i.e., the recognition of an object indicated by a human. Research conducted to improve the performance of indicated object recognition is divided into two main approaches: development and interactive. The development approach addresses the development of new devices or algorithms. Through human–robot interaction, the interactive approach improves the performance by decreasing the variability and the ambiguity of the references. Inspired by the findings of entrainment and entrainment inhibition, this study proposes a robot conversation strategy that utilizes the interactive approach. While entrainment is a phenomenon in which people unconsciously tend to mimic words and/or gestures of their interlocutor, entrainment inhibition is the opposite phenomenon in which people decrease the amount of information contained in their words and gestures when their interlocutor provides excess information. Based on these phenomena, we designed a robot conversation strategy that elicits clear references. We experimentally compared this strategy with the other interactive strategy in which a robot explicitly requests clarifications when a human refers to an object. We obtained the following findings: (1) The proposed strategy clarifies human references and improves indicated object recognition performance, and (2) the proposed strategy forms better impressions than the other interactive strategy that explicitly requests clarifications when people refer to objects.


2020 ◽  
Vol 12 (2) ◽  
pp. 230-234
Author(s):  
Junbiao Xu ◽  
Qiang Liu ◽  
Zhang Zhang ◽  
Wen Jiang ◽  
Liwen Chong

This article proposes a detection method based on thermal imaging for lead acid-battery leakage. First of all, thermal images were obtained by scanning the lead acid-battery with an infrared camera, and the images were categorized into the two sets of train and test. Then, two methods were introduced to analyze the thermal images to determine whether there was a leakage in the battery. One method used Support Vector Machine (SVM) to train the Local Binary Pattern (LBP) texture features of the images. The other method used deep learning to detect images, and trained the obtained data by DenseNet. The results demonstrate that the two methods are accurate, and show feasibility of thermal imaging to detect lead acid-battery leakage.


2021 ◽  
Author(s):  
Rohini Goel ◽  
Avinash Sharma ◽  
Rajiv Kapoor

An efficient driver assistance system is essential to avoid mishaps. The collision between the vehicles and objects before vehicle is the one of the principle reason of mishaps that outcomes in terms of diminished safety and higher monetary loss. Researchers are interminably attempting to upgrade the safety means for diminishing the mishap rates. This paper proposes an accurate and proficient technique for identifying objects in front of vehicles utilizing thermal imaging framework. For this purpose, image dataset is obtained with the help of a night vision IR camera. This strategy presents deep network based procedure for recognition of objects in thermal images. The deep network gives the model understanding of real world objects and empowers the object recognition. The real time thermal image database is utilized for the training and validation of deep network. In this work, Faster R-CNN is used to adequately identify objects in real time thermal images. This work can be an incredible help for driver assistance framework. The outcomes exhibits that the proposed work assists to boost public safety with good accuracy.


2020 ◽  
Vol 13 (2) ◽  
pp. 1
Author(s):  
E. M. Samogim ◽  
T. C. Oliveira ◽  
Z. N. Figueiredo ◽  
J. M. B. Vanini

The combine harvest for soybean crops market are currently available two types of combine with header or platform, one of conventional with revolving reel with metal or plastic teeth to cause the cut crop to fall into the auger header and the other called "draper" headers that use a fabric or rubber apron instead of a cross auger, there are few test about performance of this combine header for soybean in Mato Grosso State. The aim of this work was to evaluate the soybean harvesting quantitative losses and performance using two types combine header in four travel speed. The experiment was conducted during soybean crops season 2014/15, the farm Tamboril in the municipality of Pontes e Lacerda, State of Mato Grosso. The was used the experimental design of randomized blocks, evaluating four forward harvesting speeds (4 km h-1, 5 km h-1, 6 km h-1 and 7 km h-1), the natural crops losses were analyzed, loss caused by the combine harvester (combine header, internal mechanisms and total losses) and was also estimated the  field performance of each combine. Data were submitted to analysis of variance by F test and compared of the average by Tukey test at 5% probability. The results show the draper header presents a smaller amount of total loss and in most crop yield when compared with the conventional cross auger.


Author(s):  
D. T. Gauld ◽  
J. E. G. Raymont

The respiratory rates of three species of planktonic copepods, Acartia clausi, Centropages hamatus and Temora longicornis, were measured at four different temperatures.The relationship between respiratory rate and temperature was found to be similar to that previously found for Calanus, although the slope of the curves differed in the different species.The observations on Centropages at 13 and 170 C. can be divided into two groups and it is suggested that the differences are due to the use of copepods from two different generations.The relationship between the respiratory rates and lengths of Acartia and Centropages agreed very well with that previously found for other species. That for Temora was rather different: the difference is probably due to the distinct difference in the shape of the body of Temora from those of the other species.The application of these measurements to estimates of the food requirements of the copepods is discussed.


Sensors ◽  
2021 ◽  
Vol 21 (4) ◽  
pp. 1461
Author(s):  
Shun-Hsin Yu ◽  
Jen-Shuo Chang ◽  
Chia-Hung Dylan Tsai

This paper proposes an object classification method using a flexion glove and machine learning. The classification is performed based on the information obtained from a single grasp on a target object. The flexion glove is developed with five flex sensors mounted on five finger sleeves, and is used for measuring the flexion of individual fingers while grasping an object. Flexion signals are divided into three phases, and they are the phases of picking, holding and releasing, respectively. Grasping features are extracted from the phase of holding for training the support vector machine. Two sets of objects are prepared for the classification test. One is printed-object set and the other is daily-life object set. The printed-object set is for investigating the patterns of grasping with specified shape and size, while the daily-life object set includes nine objects randomly chosen from daily life for demonstrating that the proposed method can be used to identify a wide range of objects. According to the results, the accuracy of the classifications are achieved 95.56% and 88.89% for the sets of printed objects and daily-life objects, respectively. A flexion glove which can perform object classification is successfully developed in this work and is aimed at potential grasp-to-see applications, such as visual impairment aid and recognition in dark space.


1984 ◽  
Vol 247 (2) ◽  
pp. R250-R256
Author(s):  
H. G. Scholubbers ◽  
W. Taylor ◽  
L. Rensing

Membrane properties of whole cells of Gonyaulax polyedra were measured by fluorescence polarization. Circadian changes of fluorescence polarization exist in exponentially growing cultures. They show an amplitude larger than that of stationary cultures, indicating that a part of the change is due to or amplified by an ongoing cell cycle. Measurements of parameters of the circadian glow rhythm were analyzed for possible correlation with the membrane data. Considerable differences (Q10 = 2.5-3.0) in fluorescence polarization were found in cultures kept at different temperatures ranging from 15 to 27.5 degrees C. The free-running period length at different temperatures, on the other hand, differed only slightly (Q10 = 0.9-1.1). Stationary cultures showed higher fluorescence polarization compared with growing cultures, whereas the free-running period lengths did not differ in cultures of various densities and growth rates. Temperature steps of different sign changed the fluorescence polarization slightly in different directions. The phase shift of 4-h pulses (-5, -9, +7 degrees C) resulted in maximal phase advances of 4, 6, and 2 h, respectively. The phasing of the phase-response curves was identical in all these experiments, a finding not to be expected if the pulses act via the measured membrane properties. Pulses of drugs that change the fluorescence polarization (e.g., chlorpromazine and lidocaine) did not or only slightly phase-shift the circadian rhythm.


Sign in / Sign up

Export Citation Format

Share Document