Photoacoustic microscopy is an in vivo imaging technology based on the photoacoustic effect.It is widely used in various biomedical studies because it can provide high-resolution images while being label-free,safe,and...Photoacoustic microscopy is an in vivo imaging technology based on the photoacoustic effect.It is widely used in various biomedical studies because it can provide high-resolution images while being label-free,safe,and harmless to biological tissue.Polygon-scanning is an effective scanning method in photoacoustic microscopy that can realize fast imaging of biological tissue with a large field of view.However,in polygon-scanning,fluctuations of the rotating motor speed and the geometric error of the rotating mirror cause image distortions,which seriously affect the photoacoustic-microscopy imaging quality.To improve the image quality of photoacoustic microscopy using polygon-scanning,an image correction method is proposed based on accurate ultrasound positioning.In this method,the photoacoustic and ultrasound imaging data of the sample are simultaneously obtained,and the angle information of each mirror used in the polygon-scanning is extracted from the ultrasonic data to correct the photoacoustic images.Experimental results show that the proposed method can significantly reduce image distortions in photoacoustic microscopy,with the image dislocation offset decreasing from 24.774 to 10.365μm.展开更多
One of the main drivers for intelligent transportation systems is safety. Adaptive cruise control, as a common solution for traffic safety, lias extended from radars to cameras. Due to high mobility of vehicles and un...One of the main drivers for intelligent transportation systems is safety. Adaptive cruise control, as a common solution for traffic safety, lias extended from radars to cameras. Due to high mobility of vehicles and unevenness of roads, the picture quality of cameras has been great challenges for camera-based adaptive cruise control. In this paper, an image distortion correction algorithm is addressed. Our method is based on optical flow technology which is normally applied in motion estimation and video compression research. We are the first to attempt to adapt it in image distortion correction. Two optical flow approaches, the Lucas-Kanade method and the Horn-Schunck method, are selected and compared. The procedure of image distortion correction using the optical flow method has been tested by both synthetic test images and camera images. The experimental results show that the Lucas-Kanade method is more suitable in the correction of image distortion.展开更多
In Single-Photon Emission Computed Tomography(SPECT),the reconstructed image has insufficient contrast,poor resolution and inaccurate volume of the tumor size due to physical degradation factors.Generally,nonstationar...In Single-Photon Emission Computed Tomography(SPECT),the reconstructed image has insufficient contrast,poor resolution and inaccurate volume of the tumor size due to physical degradation factors.Generally,nonstationary filtering of the projection or the slice is one of the strategies for correcting the resolution and therefore improving the quality of the reconstructed SPECT images.This paper presents a new 3D algorithm that enhances the quality of reconstructed thoracic SPECT images and reduces the noise level with the best degree of accuracy.The suggested algorithm is composed of three steps.The first one consists of denoising the acquired projections using the benefits of the complementary properties of both the Curvelet transformand theWavelet transforms to provide the best noise reduction.The second step is a simultaneous reconstruction of the axial slices using the 3D Ordered Subset Expectation Maximization(OSEM)algorithm.The last step is post-processing of the reconstructed axial slices using one of the newest anisotropic diffusion models named Partial Differential Equation(PDE).The method is tested on two digital phantoms and clinical bone SPECT images.A comparative study with four algorithms reviewed on state of the art proves the significance of the proposed method.In simulated data,experimental results show that the plot profile of the proposed model keeps close to the original one compared to the other algorithms.Furthermore,it presents a notable gain in terms of contrast to noise ratio(CNR)and execution time.The proposed model shows better results in the computation of contrast metric with a value of 0.68±7.2 and the highest signal to noise ratio(SNR)with a value of 78.56±6.4 in real data.The experimental results prove that the proposed algorithm is more accurate and robust in reconstructing SPECT images than the other algorithms.It could be considered a valuable candidate to correct the resolution of bone in the SPECT images.展开更多
It well known that vehicle detection is an important component of the field of object detection.However,the environment of vehicle detection is particularly sophisticated in practical processes.It is comparatively dif...It well known that vehicle detection is an important component of the field of object detection.However,the environment of vehicle detection is particularly sophisticated in practical processes.It is comparatively difficult to detect vehicles of various scales in traffic scene images,because the vehicles partially obscured by green belts,roadblocks or other vehicles,as well as influence of some low illumination weather.In this paper,we present a model based on Faster ReCNN with NAS optimization and feature enrichment to realize the effective detection of multi-scale vehicle targets in traffic scenes.First,we proposed a Retinex-based image adaptive correction algorithm(RIAC)to enhance the traffic images in the dataset to reduce the influence of shadow and illumination,and improve the image quality.Second,in order to improve the feature expression of the backbone network,we conducted Neural Architecture Search(NAS)on the backbone network used for feature extraction of Faster ReCNN to generate the optimal cross-layer connection to extract multi-layer features more effectively.Third,we used the object Feature Enrichment that combines the multi-layer feature information and the context information of the last layer after cross-layer connection to enrich the information of vehicle targets,and improve the robustness of the model for challenging targets such as small scale and severe occlusion.In the implementation of the model,K-means clustering algorithm was used to select the suitable anchor size for our dataset to improve the convergence speed of the model.Our model has been trained and tested on the UN-DETRAC dataset,and the obtained results indicate that our method has art-of-state detection performance.展开更多
The shortage of current different approaches of the vehicle license plate(VLP) tilt correction is analyzed in the paper and a new rotary correction method put forward based on the former ways of the VLP tilt correctio...The shortage of current different approaches of the vehicle license plate(VLP) tilt correction is analyzed in the paper and a new rotary correction method put forward based on the former ways of the VLP tilt correction in the horizontal direction and the vertical direction Owing to the VLP tilt taking place in the vertical direction,the array of the image’s pixels of the same column is broken,and even different rows come into being superposition.The VLP tilt taking place in the horizontal direction,by which the array of the image’s pixels of the same row broken,and so much as different columns come into being superposition.展开更多
Accurate three-dimensional (3D) target positioning is of great importance in many industrial applications. Although various methods for reconstructing 3D information from a set of images have been available in the l...Accurate three-dimensional (3D) target positioning is of great importance in many industrial applications. Although various methods for reconstructing 3D information from a set of images have been available in the literature, few of them pay enough attention to the indispensable procedures, such as target extraction from images and image correction having strong influences upon the 3D positioning accuracy. This article puts forward a high-precision ellipse center (target point) extraction method and a new image correction approach which has been integrated into the 3D reconstruction pipeline with a concise implicit model to accurately compensates for the image distortion. The methods are applied to a copyright-reserved close range photogrammetric system. Real measuring experiments and industrial applications have evidenced the proposed methods, which can significantly improve the 3D positioning accuracy.展开更多
There are various applied electro-optical devices, which utilize light emitting didoe(LED) chip array for applications to displays and opto-electronic sensors. In those devices, it is the one of the critical technical...There are various applied electro-optical devices, which utilize light emitting didoe(LED) chip array for applications to displays and opto-electronic sensors. In those devices, it is the one of the critical technical issues to minimize uncertain fluctuations including optical power and optical density. Due to variation in operating environment of a device, those are not corrected precisely by controlling parameters based on simple relation between parameters and resultant abovementioned outputs.Therefore, there is essential need to correct outputs in real-time based on correction function generated from the consideration on various operation condition. In this article, we introduce an output correction method through reporting real-time image noise reduction in the application to electro-photography with LED print head. In the technology of LED print head, as differences in optical characteristics between each LED cause vertical image noise, it should be corrected in order to obtain images that are comparable or better in quality compared to those produced by the conventional laser scanning method. Even though it seems that the method used to obtain uniform light power from each LED can solve this problem, it does not work well for high-resolution printing. Therefore, a scan method involving correction by a printed and scanned pattern is introduced through this work. The scan method is composed of correction patterns to minimize printing noise by its shape, the correction algorithm to calculate the optimized value and the printing algorithm to control gray levels in real-time precisely. We believe that the developed correction method upgrades the printing quality of the LPH printer better than commercial printers. The developed correction method can also be applied to various application areas that use an array-type light source such as display systems and lighting systems.展开更多
A distortion correction method for the elemental images of integral imaging(Ⅱ) by utilizing the directional diffuser is demonstrated. In the traditional Ⅱ, the distortion originating from lens aberration wraps ele...A distortion correction method for the elemental images of integral imaging(Ⅱ) by utilizing the directional diffuser is demonstrated. In the traditional Ⅱ, the distortion originating from lens aberration wraps elemental images and degrades the image quality severely. According to the theoretical analysis and experiments, it can be proved that the farther the three-dimensional image is displayed from the lens array, the more serious the distortion is. To analyze the process of eliminating lens distortion, one lens and its corresponding elemental image are separated from the traditional Ⅱ. By introducing the directional diffuser, the aperture stop of the separated optical system changes from the eye's pupil to the lens. In terms of contrast experiments, the distortion of the improved display system is corrected effectively. In the experiment, when the distance between the reconstructed image and lens array is equal to 120 mm, the largest lens distortion is decreased from 46.6% to 3.3%.展开更多
Most modern consumer-grade cameras are often equipped with a rolling shutter mechanism,which is becoming increasingly important in computer vision,robotics and autonomous driving applications.However,its temporal-dyna...Most modern consumer-grade cameras are often equipped with a rolling shutter mechanism,which is becoming increasingly important in computer vision,robotics and autonomous driving applications.However,its temporal-dynamic imaging nature leads to the rolling shutter effect that manifests as geometric distortion.Over the years,researchers have made significant progress in developing tractable rolling shutter models,optimization methods,and learning approaches,aiming to remove geometry distortion and improve visual quality.In this survey,we review the recent advances in rolling shutter cameras from two aspects of motion modeling and deep learning.To the best of our knowledge,this is the first comprehensive survey of rolling shutter cameras.In the part of rolling shutter motion modeling and optimization,the principles of various rolling shutter motion models are elaborated and their typical applications are summarized.Then,the applications of deep learning in rolling shutter based image processing are presented.Finally,we conclude this survey with discussions on future research directions.展开更多
Functional magnetic resonance imaging(fMRI)is one of the most commonly used methods in cognitive neuroscience on humans.In recent decades,fMRI has also been used in the awake monkey experiments to localize functiona...Functional magnetic resonance imaging(fMRI)is one of the most commonly used methods in cognitive neuroscience on humans.In recent decades,fMRI has also been used in the awake monkey experiments to localize functional brain areas and to compare the functional differences between human and monkey brains.Several procedures and paradigms have been developed to maintain proper head fixation and to perform motion control training.In this study,we extended the application of fMRI to awake cats without training,receiving a flickering checkerboard visual stimulus projected to a screen in front of them in a block-design paradigm.We found that body movement-induced non-rigid motion introduced artifacts into the functional scans,especially those around the eye and neck.To correct for these artifacts,we developed two methods:one for general experimental design,and the other for studies of whether a checkerboard task could be used as a localizer to optimize the motioncorrection parameters.The results demonstrated that,with proper animal fixation and motion correction procedures,it is possible to perform fMRI experiments with untrained awake cats.展开更多
This Letter proposes a coordinate difference homogenization matching method to solve motion influence in three-dimensional(3D) range-intensity correlation laser imaging. Firstly, features and feature pairs of gate i...This Letter proposes a coordinate difference homogenization matching method to solve motion influence in three-dimensional(3D) range-intensity correlation laser imaging. Firstly, features and feature pairs of gate images are obtained by speeded-up robust figures and bi-directional feature matching methods. The original mean value of the feature-pair coordinate differences is calculated. Comparing the coordinate differences with the original mean value, the wrong feature pairs are removed, and then an optimized mean value is updated. The final feature-pair coordinates are re-registered based on the updated mean value. Thus, an accurate transformation is established to rectify motion gate images for 3D reconstruction. In the experiment, a 3D image of a tower at 780 m is successfully captured by our laser gated imaging system on a pan-tilt device.展开更多
基金This work was supported by the National Natural Science Foundation of ChinaNos.91739117 and 81927807+3 种基金Shenzhen Science and Technology Innovation,No.JCYJ20170413153129570Chinese Academy of Sciences Nos.YJKYYQ20190078 and GJJSTD20180002Shenzhen Key Laboratory for Molecular Imaging,No.ZDSY20130401165820357Guangdong Provincial Key Laboratory of Biomedical Optical Imaging,No.2020B121201010.
文摘Photoacoustic microscopy is an in vivo imaging technology based on the photoacoustic effect.It is widely used in various biomedical studies because it can provide high-resolution images while being label-free,safe,and harmless to biological tissue.Polygon-scanning is an effective scanning method in photoacoustic microscopy that can realize fast imaging of biological tissue with a large field of view.However,in polygon-scanning,fluctuations of the rotating motor speed and the geometric error of the rotating mirror cause image distortions,which seriously affect the photoacoustic-microscopy imaging quality.To improve the image quality of photoacoustic microscopy using polygon-scanning,an image correction method is proposed based on accurate ultrasound positioning.In this method,the photoacoustic and ultrasound imaging data of the sample are simultaneously obtained,and the angle information of each mirror used in the polygon-scanning is extracted from the ultrasonic data to correct the photoacoustic images.Experimental results show that the proposed method can significantly reduce image distortions in photoacoustic microscopy,with the image dislocation offset decreasing from 24.774 to 10.365μm.
文摘One of the main drivers for intelligent transportation systems is safety. Adaptive cruise control, as a common solution for traffic safety, lias extended from radars to cameras. Due to high mobility of vehicles and unevenness of roads, the picture quality of cameras has been great challenges for camera-based adaptive cruise control. In this paper, an image distortion correction algorithm is addressed. Our method is based on optical flow technology which is normally applied in motion estimation and video compression research. We are the first to attempt to adapt it in image distortion correction. Two optical flow approaches, the Lucas-Kanade method and the Horn-Schunck method, are selected and compared. The procedure of image distortion correction using the optical flow method has been tested by both synthetic test images and camera images. The experimental results show that the Lucas-Kanade method is more suitable in the correction of image distortion.
文摘In Single-Photon Emission Computed Tomography(SPECT),the reconstructed image has insufficient contrast,poor resolution and inaccurate volume of the tumor size due to physical degradation factors.Generally,nonstationary filtering of the projection or the slice is one of the strategies for correcting the resolution and therefore improving the quality of the reconstructed SPECT images.This paper presents a new 3D algorithm that enhances the quality of reconstructed thoracic SPECT images and reduces the noise level with the best degree of accuracy.The suggested algorithm is composed of three steps.The first one consists of denoising the acquired projections using the benefits of the complementary properties of both the Curvelet transformand theWavelet transforms to provide the best noise reduction.The second step is a simultaneous reconstruction of the axial slices using the 3D Ordered Subset Expectation Maximization(OSEM)algorithm.The last step is post-processing of the reconstructed axial slices using one of the newest anisotropic diffusion models named Partial Differential Equation(PDE).The method is tested on two digital phantoms and clinical bone SPECT images.A comparative study with four algorithms reviewed on state of the art proves the significance of the proposed method.In simulated data,experimental results show that the plot profile of the proposed model keeps close to the original one compared to the other algorithms.Furthermore,it presents a notable gain in terms of contrast to noise ratio(CNR)and execution time.The proposed model shows better results in the computation of contrast metric with a value of 0.68±7.2 and the highest signal to noise ratio(SNR)with a value of 78.56±6.4 in real data.The experimental results prove that the proposed algorithm is more accurate and robust in reconstructing SPECT images than the other algorithms.It could be considered a valuable candidate to correct the resolution of bone in the SPECT images.
基金This research was funded by the National Natural Science Foundation of China(grant number:61671470)the Key Research and Development Program of China(grant number:2016YFC0802900).
文摘It well known that vehicle detection is an important component of the field of object detection.However,the environment of vehicle detection is particularly sophisticated in practical processes.It is comparatively difficult to detect vehicles of various scales in traffic scene images,because the vehicles partially obscured by green belts,roadblocks or other vehicles,as well as influence of some low illumination weather.In this paper,we present a model based on Faster ReCNN with NAS optimization and feature enrichment to realize the effective detection of multi-scale vehicle targets in traffic scenes.First,we proposed a Retinex-based image adaptive correction algorithm(RIAC)to enhance the traffic images in the dataset to reduce the influence of shadow and illumination,and improve the image quality.Second,in order to improve the feature expression of the backbone network,we conducted Neural Architecture Search(NAS)on the backbone network used for feature extraction of Faster ReCNN to generate the optimal cross-layer connection to extract multi-layer features more effectively.Third,we used the object Feature Enrichment that combines the multi-layer feature information and the context information of the last layer after cross-layer connection to enrich the information of vehicle targets,and improve the robustness of the model for challenging targets such as small scale and severe occlusion.In the implementation of the model,K-means clustering algorithm was used to select the suitable anchor size for our dataset to improve the convergence speed of the model.Our model has been trained and tested on the UN-DETRAC dataset,and the obtained results indicate that our method has art-of-state detection performance.
文摘The shortage of current different approaches of the vehicle license plate(VLP) tilt correction is analyzed in the paper and a new rotary correction method put forward based on the former ways of the VLP tilt correction in the horizontal direction and the vertical direction Owing to the VLP tilt taking place in the vertical direction,the array of the image’s pixels of the same column is broken,and even different rows come into being superposition.The VLP tilt taking place in the horizontal direction,by which the array of the image’s pixels of the same row broken,and so much as different columns come into being superposition.
基金National Natural Science Foundation of China (50875 130) Doctoral Discipline Foundation of China (200802870016) Science Foundation of Jiangsu, China (BE2008136)
文摘Accurate three-dimensional (3D) target positioning is of great importance in many industrial applications. Although various methods for reconstructing 3D information from a set of images have been available in the literature, few of them pay enough attention to the indispensable procedures, such as target extraction from images and image correction having strong influences upon the 3D positioning accuracy. This article puts forward a high-precision ellipse center (target point) extraction method and a new image correction approach which has been integrated into the 3D reconstruction pipeline with a concise implicit model to accurately compensates for the image distortion. The methods are applied to a copyright-reserved close range photogrammetric system. Real measuring experiments and industrial applications have evidenced the proposed methods, which can significantly improve the 3D positioning accuracy.
基金supported by the National Research Foundation of Korea Grant funded by the Korean Government(Grant No.2015R1C1A1A01053888)the Yeungnam University Research Grant(Grant No.216A580022)
文摘There are various applied electro-optical devices, which utilize light emitting didoe(LED) chip array for applications to displays and opto-electronic sensors. In those devices, it is the one of the critical technical issues to minimize uncertain fluctuations including optical power and optical density. Due to variation in operating environment of a device, those are not corrected precisely by controlling parameters based on simple relation between parameters and resultant abovementioned outputs.Therefore, there is essential need to correct outputs in real-time based on correction function generated from the consideration on various operation condition. In this article, we introduce an output correction method through reporting real-time image noise reduction in the application to electro-photography with LED print head. In the technology of LED print head, as differences in optical characteristics between each LED cause vertical image noise, it should be corrected in order to obtain images that are comparable or better in quality compared to those produced by the conventional laser scanning method. Even though it seems that the method used to obtain uniform light power from each LED can solve this problem, it does not work well for high-resolution printing. Therefore, a scan method involving correction by a printed and scanned pattern is introduced through this work. The scan method is composed of correction patterns to minimize printing noise by its shape, the correction algorithm to calculate the optimized value and the printing algorithm to control gray levels in real-time precisely. We believe that the developed correction method upgrades the printing quality of the LPH printer better than commercial printers. The developed correction method can also be applied to various application areas that use an array-type light source such as display systems and lighting systems.
基金supported by the Natural National Science Foundation of China(NSFC)(No.61705014)the Fundamental Research Funds for the Central Universities(No.2016ZX01)+1 种基金the fund of the State Key Laboratory of Information Photonics and Optical CommunicationsThe "863" Program(No.2015AA015902)
文摘A distortion correction method for the elemental images of integral imaging(Ⅱ) by utilizing the directional diffuser is demonstrated. In the traditional Ⅱ, the distortion originating from lens aberration wraps elemental images and degrades the image quality severely. According to the theoretical analysis and experiments, it can be proved that the farther the three-dimensional image is displayed from the lens array, the more serious the distortion is. To analyze the process of eliminating lens distortion, one lens and its corresponding elemental image are separated from the traditional Ⅱ. By introducing the directional diffuser, the aperture stop of the separated optical system changes from the eye's pupil to the lens. In terms of contrast experiments, the distortion of the improved display system is corrected effectively. In the experiment, when the distance between the reconstructed image and lens array is equal to 120 mm, the largest lens distortion is decreased from 46.6% to 3.3%.
基金This work was supported in part by National Natural Science Foundation of China(Nos.62271410,61901387 and 62001394)the Fundamental Research Funds for the Central Universities,China,and the Innovation Foundation for Doctor Dissertation of Northwestern Polytechnical University,China(No.CX2022046).
文摘Most modern consumer-grade cameras are often equipped with a rolling shutter mechanism,which is becoming increasingly important in computer vision,robotics and autonomous driving applications.However,its temporal-dynamic imaging nature leads to the rolling shutter effect that manifests as geometric distortion.Over the years,researchers have made significant progress in developing tractable rolling shutter models,optimization methods,and learning approaches,aiming to remove geometry distortion and improve visual quality.In this survey,we review the recent advances in rolling shutter cameras from two aspects of motion modeling and deep learning.To the best of our knowledge,this is the first comprehensive survey of rolling shutter cameras.In the part of rolling shutter motion modeling and optimization,the principles of various rolling shutter motion models are elaborated and their typical applications are summarized.Then,the applications of deep learning in rolling shutter based image processing are presented.Finally,we conclude this survey with discussions on future research directions.
基金supported by grants from the Ministry of Science and Technology of China (2012CB825500, 2012IM030100, 2010IM030800)the National Natural Science Foundation of China (91132302, 90820307)
文摘Functional magnetic resonance imaging(fMRI)is one of the most commonly used methods in cognitive neuroscience on humans.In recent decades,fMRI has also been used in the awake monkey experiments to localize functional brain areas and to compare the functional differences between human and monkey brains.Several procedures and paradigms have been developed to maintain proper head fixation and to perform motion control training.In this study,we extended the application of fMRI to awake cats without training,receiving a flickering checkerboard visual stimulus projected to a screen in front of them in a block-design paradigm.We found that body movement-induced non-rigid motion introduced artifacts into the functional scans,especially those around the eye and neck.To correct for these artifacts,we developed two methods:one for general experimental design,and the other for studies of whether a checkerboard task could be used as a localizer to optimize the motioncorrection parameters.The results demonstrated that,with proper animal fixation and motion correction procedures,it is possible to perform fMRI experiments with untrained awake cats.
基金supported by the National Key Research and Development Program of China(No.2016YFC0500103)the Youth Innovation Promotion Association CAS(No.2017155)the Scientific Instrument Development Project from Capital Science and Technology Condition Platform(No.Z171100002817002)
文摘This Letter proposes a coordinate difference homogenization matching method to solve motion influence in three-dimensional(3D) range-intensity correlation laser imaging. Firstly, features and feature pairs of gate images are obtained by speeded-up robust figures and bi-directional feature matching methods. The original mean value of the feature-pair coordinate differences is calculated. Comparing the coordinate differences with the original mean value, the wrong feature pairs are removed, and then an optimized mean value is updated. The final feature-pair coordinates are re-registered based on the updated mean value. Thus, an accurate transformation is established to rectify motion gate images for 3D reconstruction. In the experiment, a 3D image of a tower at 780 m is successfully captured by our laser gated imaging system on a pan-tilt device.