An Efficient Bar Code Image Recognition Algorithm for Sorting System

In the sorting system of the production line, the object movement, fixed angle of view, light intensity and other reasons lead to obscure blurred images. It results in bar code recognition rate being low and real time being poor. Aiming at the above problems, a progressive bar code compressed recognition algorithm is proposed. First, assuming that the source image is not tilted, use the direct recognition method to quickly identify the compressed source image. Failure indicates that the compression ratio is improper or the image is skewed. Then, the source image is enhanced to identify the source image directly. Finally, the inclination of the compressed image is detected by the barcode region recognition method and the source image is corrected to locate the barcode information in the barcode region recognition image. The results of multitype image experiments show that the proposed method is improved by 5+ times computational efficiency compared with the former methods, and can recognize fuzzy images better.     

Polarimetric imaging method for target enhancement in haze based on polarimetric retrieval

Polarimetric imaging has been proven to be an effective way in detecting the targets of interest in complicated surroundings by analyzing the polarization property, instead of the intensity, of the light emanating from the objects. Unfortunately, polarimetric imaging encounters difficulty when the surroundings are very scattered, where on the one hand the polarization property of the object light usually becomes very faint after a strong depolarized scattering process; on the other hand, the object light will be blended with the atmospheric light scattered by haze particles (airlight). In this paper, we propose a polarimetric imaging retrieval method that can be used for such challenging conditions. Firstly, the airlight radiance is estimated precisely. Then, the airlight is removed from the hazy images. Finally, the residual polarization property of the object light is regained, which ensures the validity of the polarimetric imaging method in these conditions. The experiments confirm that the proposed method is useful in enhancing polarimetric imaging detection in haze.     

Three-dimensional object detection under arbitrary lighting conditions

A novel method of 3D object recognition independent of lighting conditions is presented. The recognition model is based on a vector space representation using an orthonormal basis generated by the Lambertian reflectance functions obtained with distant light sources. Changing the lighting conditions corresponds to multiplying the elementary images by a constant factor and because of that, all possible lighting views will be elements that belong to that vector space. The recognition method proposed is based on the calculation of the angle between the vector associated with a certain illuminated 3D object and that subspace. We define the angle in terms of linear correlations to get shift and illumination-invariant detection.     

Multiple Pedestrian Detection and Tracking in Night Vision Surveillance Systems

Pedestrian detection and tracking are vital elements of today’s surveillance systems, which make daily life safe for humans. Thus, human detection and visualization have become essential inventions in the field of computer vision. Hence, developing a surveillance system with multiple object recognition and tracking, especially in low light and night-time, is still challenging. Therefore, we propose a novel system based on machine learning and image processing to provide an efficient surveillance system for pedestrian detection and tracking at night. In particular, we propose a system that tackles a two-fold problem by detecting multiple pedestrians in infrared (IR) images using machine learning and tracking them using particle filters. Moreover, a random forest classifier is adopted for image segmentation to identify pedestrians in an image. The result of detection is investigated by particle filter to solve pedestrian tracking. Through the extensive experiment, our system shows 93% segmentation accuracy using a random forest algorithm that demonstrates high accuracy for background and roof classes. Moreover, the system achieved a detection accuracy of 90% using multiple template matching techniques and 81% accuracy for pedestrian tracking. Furthermore, our system can identify that the detected object is a human. Hence, our system provided the best results compared to the state-of-art systems, which proves the effectiveness of the techniques used for image segmentation, classification, and tracking. The presented method is applicable for human detection/tracking, crowd analysis, and monitoring pedestrians in IR video surveillance.     

Towards time-efficient ghost imaging

Reducing the time necessary to acquire information is highly desirable in almost every context. Ghost imaging is no exception, which is very time consuming due to its scanning nature and low light levels innate to quantum experiments. This work aimed to reduce the time required to reconstruct the image whilst maintaining quality. In doing so, we followed two complementary approaches: one varying the experimental parameters, and another implementing computational processing. We defined a performance measure based on the image reconstruction time and its resemblance to the original object, and determined that the use of image processing and recognition algorithms offers major improvements in temporal efficiency. Importantly, if the main purpose of imaging is solely object recognition, low resolution mask patterns give better results, whereas higher resolution patterns yield better resolved images, at the expense of time. We believe this work will pique interest in the ghost and single-pixel imaging communities.     

动态图像的拼接与运动目标检测方法的研究

动态图像运动目标检测是图像处理中的热点,但动态图像的识别范围却成了目标检测的限制,针对此问题,本文提出了一种利用图像拼接技术扩展图像识别范围、并在此基础上完成运动目标检测的方法。在图像拼接中采用了SURF图像匹配算法,运动目标识别利用背景差分法,实验中使用的是开源的Linux操作系统、以及为图像处理提供了大量算法和函数的Open CV软件开发库。针对不同分辨率、不同角度采集的图像进行了实验研究,结果表明,可以在较好满足图像识别范围的同时,明确地检测出运动目标的相关信息。同时,本文提出一种通过图像拼接实现扩展运动目标检测的方法,满足了实时性要求,达到了增加图像清晰度的目的,但是,在摄像设备与场景之间的相对运动方面还存在着有待解决的问题,这将成为今后研究的重点方向。     

Improvement of face recognition using light compensation technique on real-time imaging

Camera imaging systems are used widely. However, the resulting images may show unequal light distributions due to backlight. In this paper, an adaptive backlight compensation algorithm is presented for fixing the brightness and contrast in regions of interest, particularly for human faces. The framework is implemented in two stages. The first stage is the light compensation algorithm, which depends on face detection and focuses on the intensities of pixels in ‘face’ regions only. The second stage uses a distance weighting approach to address artificial light effects created by the first stage. This algorithm can adjust the imaging light distribution adaptively in order to solve the problem of backlight and achieve natural-looking pictures. For face recognition systems, this approach can improve the success rate for face recognition by 35% on average when images are backlit.     

Modified gamma correction method to enhance ball grid array image for surface defect inspection

BGA (Ball Grid Array) surface defect detection requires faster and more accurate methods for semiconductor industry applications. Traditionally, the BGA inspection used gray-scale images. However, the solder pad, wiring and gray scales shown in images depict little variance. Therefore, when the threshold value is poorly set or the contract rate is insignificant, BGA detection may fail to segment an object. This research proposes a modified methodology that uses Gamma correction for image enhancement. Three-color bands were applied to a modified Gamma correction algorithm (i.e. RGB) to better separate the high and low image contrasts. Better results were obtained by dividing the image into background and foreground portions using the Gamma correction. As a result, the proposed method improved the contrast value by 52.09%. After the images were enhanced and segmented, the compactness and internal holes were calculated as features for classification. The results showed that classification correctness was 96.43%. The proposed method used a 640?×?480 pixel image, performing complete defect detection 0.3 seconds faster than the traditional enhancement method, which requires 1?second. The research results provide an effective solution for the detection and classification of the BGA surface tin ball defect problem.     

基于NSST多尺度自适应的Retinex低照度图像增强算法

目的在对低照度图像进行增强时,针对传统频率域方法由于尺度不够丰富而不能很好保留图像高频细节的问题,提出一种基于NSST多尺度自适应的Retinex低照度图像增强算法。方法首先将低照度图像转化至HSI颜色空间后,单独对I通道进行处理,实现对图像色彩信息的保真效果;然后对I通道进行Retinex算法得到反射分量,从而去除照度信息对图像亮度的影响;对反射分量进行伽马调整后,进行基于La(平均亮度)、Pa(平均对比度)、Ia(信息熵)等3个特征值的自适应NSST分解,从而得到最佳参数的高频分量。结果在主观观察和客观无参考图像质量评价中,文中算法的增强效果和评价得分都要优于其他算法。结论经过自适应参数优化之后,低照度图像的对比度得到了提高,可视性和图像质量都得到了显著提升。     

Occlusion invariant face recognition using mean based weight matrix and support vector machine

In this paper, a novel occlusion invariant face recognition algorithm based on Mean based weight matrix (MBWM) technique is proposed. The proposed algorithm is composed of two phases—the occlusion detection phase and the MBWM based face recognition phase. A feature based approach is used to effectively detect partial occlusions for a given input face image. The input face image is first divided into a finite number of disjointed local patches, and features are extracted for each patch, and the occlusion present is detected. Features obtained from the corresponding occlusion-free patches of training images are used for face image recognition. The SVM classifier is used for occlusion detection for each patch. In the recognition phase, the MBWM bases of occlusion-free image patches are used for face recognition. Euclidean nearest neighbour rule is applied for the matching. GTAV face database that includes many occluded face images by sunglasses and hand are used for the experiment. The experimental results demonstrate that the proposed local patch-based occlusion detection technique works well and the MBWM based method shows superior performance to other conventional approaches.     

一种声图像局部自适应增强方法

声图像具有分辨率较低、混响干扰强、物体区域面积小、边界轮廓模糊不清等特点，因此采用全局的增强算法不能取得满意的结果。根据声图像特点，定义了局部增强函数，以边缘数目、边缘强度和熵定义了评价函数，通过采用PSO算法对增强函数中的参数进行了优化选取。最后与已实现的迭代自适应增强算法和直方图均衡化算法进行了对比分析，结果表明，该算法在保持了原图的灰度级的基础上，不但明显改善图像视觉效果，而且有效地增强和保留图像细节，改善图像质量，利于后续图像检测，是一种适合于声图像的增强方法。     

Spherical nonlinear correlations for global invariant three-dimensional object recognition

We define a nonlinear filtering based on correlations on unit spheres to obtain both rotation- and scale-invariant three-dimensional (3D) object detection. Tridimensionality is expressed in terms of range images. The phase Fourier transform (PhFT) of a range image provides information about the orientations of the 3D object surfaces. When the object is sequentially rotated, the amplitudes of the different PhFTs form a unit radius sphere. On the other hand, a scale change is equivalent to a multiplication of the amplitude of the PhFT by a constant factor. The effect of both rotation and scale changes for 3D objects means a change in the intensity of the unit radius sphere. We define a 3D filtering based on nonlinear operations between spherical correlations to achieve both scale- and rotation-invariant 3D object recognition.     

Text Detection and Recognition for Natural Scene Images Using Deep Convolutional Neural Networks

Words are the most indispensable information in human life. It is very important to analyze and understand the meaning of words. Compared with the general visual elements, the text conveys rich and high-level moral information, which enables the computer to better understand the semantic content of the text. With the rapid development of computer technology, great achievements have been made in text information detection and recognition. However, when dealing with text characters in natural scene images, there are still some limitations in the detection and recognition of natural scene images. Because natural scene image has more interference and complexity than text, these factors make the detection and recognition of natural scene image text face many challenges. To solve this problem, a new text detection and recognition method based on depth convolution neural network is proposed for natural scene image in this paper. In text detection, this method obtains high-level visual features from the bottom pixels by ResNet network, and extracts the context features from character sequences by BLSTM layer, then introduce to the idea of faster R-CNN vertical anchor point to find the bounding box of the detected text, which effectively improves the effect of text object detection. In addition, in text recognition task, DenseNet model is used to construct character recognition based on Kares. Finally, the output of Softmax is used to classify each character. Our method can replace the artificially defined features with automatic learning and context-based features. It improves the efficiency and accuracy of recognition, and realizes text detection and recognition of natural scene images. And on the PAC2018 competition platform, the experimental results have achieved good results.     

Image contrast enhancement with brightness preservation using an optimal gamma correction and weighted sum approach

The enhancement of image contrast and preservation of image brightness are two important but conflicting objectives in image restoration. Previous attempts based on linear histogram equalization had achieved contrast enhancement, but exact preservation of brightness was not accomplished. A new perspective is taken here to provide balanced performance of contrast enhancement and brightness preservation simultaneously by casting the quest of such solution to an optimization problem. Specifically, the non-linear gamma correction method is adopted to enhance the contrast, while a weighted sum approach is employed for brightness preservation. In addition, the efficient golden search algorithm is exploited to determine the required optimal parameters to produce the enhanced images. Experiments are conducted on natural colour images captured under various indoor, outdoor and illumination conditions. Results have shown that the proposed method outperforms currently available methods in contrast to enhancement and brightness preservation.     

用于室外道路环境综合理解的多种新型视觉传感技术和系统

针对室外智能移动机器人自主导航的要求，提出了基于多种新型视觉传感技术的室外道路环境综合理解方法，其基本原理是综合利用机器人四周360°景物的环视图象信息、机器人前方道路的双目注视图象信息以及机器人运行过程中形成的时空全景图象信息，综合完成实时机器人行驶方向确定、实时路面障碍物检测和机器人全局定位等视觉任务。     

Geometric centre tracking of tagged objects using a low power demodulation smart vision sensor

In this study, a modulated light detecting smart CMOS image sensor is presented. The design has the ability to sense asynchronous signals transmitted from electronic markers such as flashing light emitting diodes (LEDs) tagged on moving objects. The geometric centre of the detected region is returned as the output result. With the presented sensor, object localisation and position detection functions are simplified, performed at higher speeds in real time and power requirement is reduced. The sensor in-pixel processing filters out the background image data, detects the modulated marker regions and projects the extracted region on the two axes, while the geometric centre extraction units placed at each axis identify the coordinates assigned to the marker. The design presents less sensitivity to object texture compared with techniques based on edge extraction or binarisation. The sensor has been designed as a 64 x 64 pixel VLSI CMOS chip in the 0.35 μm standard CMOS technology and analysed in the presence of mismatches and noise. Issues such as sensor array scalability, speed and power dissipation are also examined in this study and features of the sensor are reported and compared with some previous designs.     

基于图像识别算法的图像传感器质量检测方法改善研究

针对图像传感器质量检测结果可信度低的问题，基于Laplacian算子图像识别技术，借助Matlab软件，建立图像识别模型，实现图像采集、图像边缘智能识别、图像特征信息智能提取与分析的功能。通过实验对比，证明改善后的检测方法能更准确地反映产品质量特性。     

Optical pattern recognition using a unidirectional photorefractive oscillator coupled with an angular multiplexing volume hologram

Abstract

We propose and demonstrate a unidirectional photorefractive ring oscillator that couples with an external angular multiplexing volume hologram for pattern recognition. By configuring the hologram externally, a computer-generated hologram (CGH) and a spatial light modulator are utilized to generate reference beams for the hologram. Two-dimensional images are stored in the form of resonating beams in the system, whereby the nonlinear interaction between the beams allows the image that most resembles an input object to be recognized. Five images were used to illustrate the pattern recognition ability of the oscillator. It was found that the input object was successfully recognized within 10s. The fact that CGHs can generate a large number of beamlets enhances significantly the storage capacity in this system.     

Real-time optical image subtraction and edge enhancement using ferroelectric liquid-crystal devices based on speckle modulation

We carried out real-time optical image subtraction and edge enhancement based on a speckle modulation technique by using ferroelectric liquid-crystal polarization switches and a ferroelectric liquid-crystal spatial light modulator. A ferroelectric liquid-crystal spatial light modulator is employed as a real-time and multiple-exposure optical device, and successful results are obtained from three-exposure images modulated by speckles. Thus, image subtraction and edge enhancement are realized in real time. The whole operation is performed within several milliseconds with modest operating conditions. Because the spatial light modulator has a high resolution of greater than 100 line pairs/mm and can store fine speckle patterns, the image qualities we obtained are quite satisfactory.