Traditional Object Recognition Research Articles

Learning-based robotic grasping: A review.

As personalization technology increasingly orchestrates individualized shopping or marketing experiences in industries such as logistics, fast-moving consumer goods, and food delivery, these sectors require flexible solutions that can automate object grasping for unknown or unseen objects without much modification or downtime. Most solutions in the market are based on traditional object recognition and are, therefore, not suitable for grasping unknown objects with varying shapes and textures. Adequate learning policies enable robotic grasping to accommodate high-mix and low-volume manufacturing scenarios. In this paper, we review the recent development of learning-based robotic grasping techniques from a corpus of over 150 papers. In addition to addressing the current achievements from researchers all over the world, we also point out the gaps and challenges faced in AI-enabled grasping, which hinder robotization in the aforementioned industries. In addition to 3D object segmentation and learning-based grasping benchmarks, we have also performed a comprehensive market survey regarding tactile sensors and robot skin. Furthermore, we reviewed the latest literature on how sensor feedback can be trained by a learning model to provide valid inputs for grasping stability. Finally, learning-based soft gripping is evaluated as soft grippers can accommodate objects of various sizes and shapes and can even handle fragile objects. In general, robotic grasping can achieve higher flexibility and adaptability, when equipped with learning algorithms.

Traffic sign recognition based on deep learning

AbstractIntelligent Transportation System (ITS), including unmanned vehicles, has been gradually matured despite on road. How to eliminate the interference due to various environmental factors, carry out accurate and efficient traffic sign detection and recognition, is a key technical problem. However, traditional visual object recognition mainly relies on visual feature extraction, e.g., color and edge, which has limitations. Convolutional neural network (CNN) was designed for visual object recognition based on deep learning, which has successfully overcome the shortcomings of conventional object recognition. In this paper, we implement an experiment to evaluate the performance of the latest version of YOLOv5 based on our dataset for Traffic Sign Recognition (TSR), which unfolds how the model for visual object recognition in deep learning is suitable for TSR through a comprehensive comparison with SSD (i.e., single shot multibox detector) as the objective of this paper. The experiments in this project utilize our own dataset. Pertaining to the experimental results, YOLOv5 achieves 97.70% in terms of mAP@0.5 for all classes, SSD obtains 90.14% mAP in the same term. Meanwhile, regarding recognition speed, YOLOv5 also outperforms SSD.

Object recognition algorithm based on optimized nonlinear activation function-global convolutional neural network

Traditional object recognition algorithms cannot meet the requirements of object recognition accuracy in the actual warehousing and logistics field. In recent years, the rapid development of the deep learning theory has provided a technical approach for solving the above problems, and a number of object recognition algorithms has been proposed based on deep learning, which have been promoted and applied. However, deep learning has the following problems in the application process of object recognition: First, the nonlinear modeling ability of the activation function in the deep learning model is poor; second, the deep learning model has a large number of repeated pooling operations during which information is lost. In view of these shortcomings, this paper proposes multiple-parameter exponential linear units with uniform and learnable parameter forms and introduces two learned parameters in the exponential linear unit (ELU), enabling it to represent piecewise linear and exponential nonlinear functions. Therefore, the ELU has good nonlinear modeling capabilities. At the same time, to improve the problem of losing information in the large number of repeated pooling operations, this paper proposes a new global convolutional neural network structure. This network structure makes full use of the local and global information of different layer feature maps in the network. It can reduce the problem of losing feature information in the large number of pooling operations. Based on the above ideas, this paper suggests an object recognition algorithm based on the optimized nonlinear activation function-global convolutional neural network. Experiments were carried out on the CIFAR100 dataset and the ImageNet dataset using the object recognition algorithm proposed in this paper. The results show that the object recognition method suggested in this paper not only has a better recognition accuracy than traditional machine learning and other deep learning models but also has a good stability and robustness.

Mirrored conditional random field model for object recognition in indoor environments

Traditional object recognition algorithms are based on a commonly adopted closed-set hypothesis, assuming that the knowledge given in training is complete. However, real situations are often open and nonstatic, in which case the models only obtain incomplete knowledge during the training phase. This paper proposes a new type of conditional random field (CRF) model to solve a special case of incomplete knowledge, in which the visual appearance of certain objects changes significantly between training and testing, and as a result, certain unary features (features of individual objects) extracted from red green blue-depth (RGB-D) images are no longer reliable. Mirror nodes are introduced into the architecture based on the standard CRF model to build the mirrored conditional random field (Mirror-CRF) model, which integrates two types of object nodes: original nodes and mirror nodes. The mirror nodes have no unary features, only pairwise features, which describe relationships between two objects and are more reliable than unary features for object recognition in the case of appearance variation. The experimental results show that the Mirror-CRF model reduces the influence of significant changes in the appearance of certain objects and improves the object recognition ability under the condition of incomplete knowledge.

Context-Dependent Object Proposal and Recognition

Accurate and fast object recognition is crucial in applications such as automatic driving and unmanned aerial vehicles. Traditional object recognition methods relying on image-wise computations cannot afford such real-time applications. Object proposal methods appear to fit into this scenario by segmenting object-like regions to be further analyzed by sophisticated recognition models. Traditional object proposal methods have the drawback of generating many proposals in order to maintain a satisfactory recall of true objects. This paper presents two proposal refinement strategies based on low-level cues and context-dependent features, respectively. The low-level cues are used to enhance the edge image, while the context-dependent features are verified to rule out false objects that are irrelevant to our application. In particular, the context of the drink commodity is considered because the drink commodity has the largest sales in Taiwan’s convenience store chains, and the analysis of its context has great value in marketing and management. We further developed a support vector machine (SVM) based on the Bag of Words (BoW) model with scale-invariant feature transform (SIFT) descriptors to recognize the proposals. The experimental results show that our object proposal method generates many fewer proposals than those generated by Selective Search and EdgeBoxes, with similar recall. For the performance of SVM, at least 82% of drink objects are correctly recognized for test datasets of various challenging difficulties.

Apple Detection in Natural Environment Using Deep Learning Algorithms

It is a challenging problem to detect the apple in natural environment using traditional object recognition algorithms due to occlusion, fluctuating illumination and complex backgrounds. Deep learning methods for object detection make impressive progress, which can automatically extract the number, pixel position, size and other features of apples from the images. In this paper, four deep learning recognition models, Faster RCNN based on AlexNet, Faster RCNN based on ResNet101, YOLOv3 based on DarkNet53 and improved YOLOv3 were employed to carry out recognition experiments on red and green apple under three illumination and two image sharpness conditions, with the transfer learning to accelerate the training process. The results showed that improved YOLOv3 model had the best recognition effect among the four detection models. F1 value of red apple recognition was 95.0%, 94.6% and 94.1% for normal, insufficient and excessive illumination, respectively, and F1 value of green apple recognition was 94.9%, 94.0% and 91.1%. There were F1 value of 92.8% and 92.1% for red and green apple recognition in blurred images, respectively. Moreover, improved YOLOv3 algorithm still had the better performance for occlusion, spot, overlap and incomplete apples, with a recognition recall rate higher than 88.5%. It can be concluded that improved YOLOv3 algorithm can provide a more efficient way for apple detection in natural environment.

Research on key technology of pavement object recognition based on machine learning

According to statistics, the incidence of traffic accidents in China is far greater than that of developed countries such as the USA and Germany. Nearly 100,000 people die in traffic accidents every year in China. The use of the assisted driving system effectively reduces the incidence of traffic accidents, while road surface object recognition is the key research object in the assisted driving system. At present, in the research on the recognition of pavement objects, there are problems such as low recognition rate and long recognition time, which cannot play the role of assisting driving. In response to these problems, this paper proposes a research on pavement object recognition based on machine learning, including analysis of key technologies of machine learning and optimization of related algorithms. In order to verify the feasibility of the proposed method in the assisted driving system, the research results are applied to the road surface object recognition device and the experiment is carried out under the actual traffic environment. The results show that in the selected experimental scenes, the recognition accuracy of the road surface object recognition device applying the research results of this paper is 100%, and the recognition time is far lower than the traditional road object recognition method. The results show that the proposed method can quickly and effectively identify the pavement object, and then assist the driver to control the vehicle, which has good applicability in the assisted driving system.

Door Recognition Principles in GITA’s Machine Vision

Nowadays, in the traditional door or object recognition system, people use GluonCV and Tensorflow to help teach machines millions of pictures, and then base on the information in the database collected from the pictures to do the door recognition. While in the new proposal, a new door recognition system based on a machine carrier called GITA is proposed. In the proposal, the recognition elements comprise motionless elements, including door boundary, handle/bar, and ratio of body width and door width, and loco-motor elements, which include people’s gesture, change of door depth and change of illumination. In order to prove it, several different tests based on seven doors of different types are done.

Integrated Localization and Recognition for Inshore Ships in Large Scene Remote Sensing Images

Automatic inshore ship recognition, which includes target localization and type recognition, is an important and challenging task. However, existing ship recognition methods mainly focus on the classification of ship samples or clips. These methods rely deeply on the detection algorithm to complete localization and recognition in large scene images. In this letter, we present an integrated framework to automatically locate and recognize inshore ships in large scene satellite images. Different from traditional object recognition methods using two steps of detection-classification, the proposed framework could locate inshore ships and identify types without the detection step. Considering ship size is a useful feature, a novel multimodel method is proposed to utilize this feature. And an Euclidean-distance-based fusion strategy is used to combine candidates given by models. This fusion strategy could effectively separate side-by-side ships. To handle large scene images efficiently, scale-invariant feature transform registration is also integrated into the framework to utilize geographic information. All of these make the framework an end-to-end fashion which could automatically recognize inshore ships in large scene satellite images. Experiments on Quickbird images show that this framework could achieve the actual applied requirements.

Recognizing multi-view objects with occlusions using a deep architecture

Image-based object recognition is employed widely in many computer vision applications such as image semantic annotation and object location. However, traditional object recognition algorithms based on the 2D features of RGB data have difficulty when objects overlap and image occlusion occurs. At present, RGB-D cameras are being used more widely and the RGB-D depth data can provide auxiliary information to address these challenges. In this study, we propose a deep learning approach for the efficient recognition of 3D objects with occlusion. First, this approach constructs a multi-view shape model based on 3D objects by using an encode–decode deep learning network to represent the features. Next, 3D object recognition in indoor scenes is performed using random forests. The application of deep learning to RGB-D data is beneficial for recovering missing information due to image occlusion. Our experimental results demonstrate that this approach can significantly improve the efficiency of feature representation and the performance of object recognition with occlusion.

Fast Recognition of Mechanical Objects Using Neural Networks Under Robust Aspect

Image recognition systems, which are invariant to rotation and scale, can be useful for a variety of automated-tasks, and, therefore, command considerable interest. A fast and highly robust vision system is very important in real-time object recognition. Neural network, which allows large parallel interconnections, presents a promising alternative to traditional real-time object recognition techniques such as template matching. In this paper, a novel neural-network based and scale and rotation invariant object recognition system is presented, that is faster than the template matching technique traditionally used, and thus, has an edge in real-time operation. Results obtained are included in the paper that compare favorably with the template-matching technique in terms of search time.

User-Centric Environment Discovery With Camera Networks in Smart Homes

We propose a data integration and reasoning technique for automatic environment discovery in smart homes based on observations of user interactions with objects. This approach is complementary to traditional appearance-based object recognition, which often demands large training sets. In our approach, object recognition is achieved in a semantic way by linking object behaviors to the pose and activity of the person using them. The complex relations between objects and user activities are modeled with a Markov logic network. The embodiments of the proposed approach in two multicamera smart environments are described, and the experimental results are presented.

OBJECT CLASSIFICATION AND OCCLUSION HANDLING USING QUADRATIC FEATURE CORRELATION MODEL AND NEURAL NETWORKS

Occlusion handling is an old but important problem for the computer vision and pattern recognition community. Features from different objects may twist with each other, and any matched feature points may belong to different objects for many traditional object recognition algorithms. To recognize occlusions, we should not only match objects from different view points but also match features extracted from the same object. In this paper, we propose a method to consider these two perspectives simultaneously by encoding various types of features, such as geometry, color and texture relationships among feature points into a matrix and find the best quadratic feature correlation model to fit them. Experiments on our own built dataset and the publicly available PASCAL VOC dataset shows that, our method can robustly classify objects and handle occluded objects under large occlusions, and the performance is among the state-of-the-art.

Line Feature Based Man-Made Object Recognition with Invariance

The traditional three-dimensional object recognition method based on hypothesis and test needs to solve the coordinate transformation matrix from scene to model through a group of non-linear equations.Therefore,it has a very high complexity.This paper presents a man-made object recognition method based on the geometric feature of line characteristics,and disperses the overall coordinate transformation calculation in every local Plane Homography calculation,reduces the complexity of the solution.Its process: firstly pre-match the feature points using geometric invariants,then assume and solve the Plane Homography matrix between scenes to model.After that it matches the line segments on the homography plane,and by this verifies the assumption.Experiments prove that this method can rapidly and accurately identify man-made objects with coplanar line features.

Real-Time Tracking of Multiple Objects Using Fiducials for Augmented Reality

Augmented reality requires understanding of the scene to know when, where and what to display as a response to changes in the surrounding world. This understanding often involves tracking and recognition of multiple objects and locations in real-time. Technologies frequently used for multiple object tracking, such as electromagnetic trackers are very limited in range, as well as constraining. The use of Computer Vision to identify and track multiple objects is very promising. However, the requirements for traditional object recognition using appearance-based or model-based vision are very complex and their performance is far from real-time. An alternative is to use a set of markers or fiducials for object tracking and recognition. In this paper we present a system of marker coding that, together with an efficient image processing technique, provides a practical method for tracking the marked objects in real-time. The technique is based on clustering of candidate regions in space using a minimum spanning tree. The markers in the codes also allow the estimation of the three dimensional pose of the objects. We demonstrate the utility of the marker-based tracking technique in an Augmented Reality application. The application involves superimposing graphics over real industrial parts that are tracked using fiducials and manipulated by a human in order to complete an assembly. The system aids in the evaluation of the different assembly sequence possibilities.

Use neural networks to determine matching order for recognizing overlapping objects

In traditional model-based object recognition systems, a model of each object in the model database is matched in a random sequence against the scene image. The matching procedure must be repeated for every model in the database until a correct match is found. The major problem with such an approach is that as the number of models is increased the computational time required to find the correct match becomes very high. In this paper, we present an artificial neural network (ANN) approach to determine the matching order of models in the database. The match between a given scene image and a model is based on the rank of similarity of the model rather than its serial storage order in the database. Both isolated and overlapping objects that comprise piecewise linear and circular segments are considered for the recognition. Experimental results have shown that the proposed ANN approach succeeds in recognizing isolated objects, and achieves significant gain, in number of matches, over traditional model-based object recognition systems for identifying overlapping objects.