Grasp Detection under Occlusions Using SIFT Features

Distinguishing target object under occlusions has become the forefront of research to cope with grasping study in general. In this paper, a novel framework which is able to be utilized for a parallel robotic gripper is proposed. There are two key steps for the proposed method in the process of grasping occluded object: generating template information and grasp detection using the matching algorithm. A neural network, trained by the RGB-D data from the Cornell Grasp Dataset, predicts multiple grasp rectangles on template images. A proposed matching algorithm is utilized to eliminate the influence caused by occluded parts on scene images and generates multiple grasp rectangles for objects under occlusions using the grasp information of matched template images. In order to improve the quality of matching result, the proposed matching algorithm improves the SIFT algorithm and combines it with the improved RANSAC algorithm. In this way, this paper obtains suitable grasp rectangles on scene images and offers a new thought about grasping detection under occlusions. The validation results show the effectiveness and efficiency of this approach.

Panoramic image mosaic method based on image segmentation and Improved SIFT algorithm

Aiming at the problem of large amount of calculation in extracting image feature points in panoramic image mosaic by SIFT algorithm, a panoramic image mosaic algorithm based on image segmentation and Improved SIFT is proposed in this paper. The algorithm fully considers the characteristics of panoramic image stitching. Firstly, the stitched image is divided into blocks, and the maximum overlapping block of image pairs is extracted by using mutual information. The SIFT key points are extracted by SIFT algorithm, and the dog is filtered before the spatial extreme value detection of SIFT algorithm to eliminate the feature points with small intensity value; When establishing the feature descriptor, the 128 dimension of the original algorithm is reduced to 64 dimensions to reduce the amount of calculation. In the feature point registration process, the feature descriptor is reduced to 32 dimensions, the feature point pairs are roughly extracted by the optimal node first BBF algorithm, and the feature point pairs are registered and screened by RANSAC; Finally, the image transformation matrix is obtained to realize panoramic image mosaic. The experimental results show that the proposed algorithm not only ensures the panoramic mosaic effect, but also extracts the feature points in 11% of the time of the traditional SIFT algorithm, and the feature point registration speed is 27.17% of the traditional SIFT algorithm.

A CNN-Based Advertisement Recommendation through Real-Time User Face Recognition

The advertising market’s use of smartphones and kiosks for non-face-to-face ordering is growing. An advertising video recommender system is needed that continuously shows advertising videos that match a user’s taste and displays other advertising videos quickly for unwanted advertisements. However, it is difficult to make a recommender system to identify users’ dynamic preferences in real time. In this study, we propose an advertising video recommendation procedure based on computer vision and deep learning, which uses changes in users’ facial expressions captured at every moment. Facial expressions represent a user’s emotions toward advertisements. We can utilize facial expressions to find a user’s dynamic preferences. For such a purpose, a CNN-based prediction model was developed to predict ratings, and a SIFT algorithm-based similarity model was developed to search for users with similar preferences in real time. To evaluate the proposed recommendation procedure, we experimented with food advertising videos. The experimental results show that the proposed procedure is superior to benchmark systems such as a random recommendation, an average rating approach, and a typical collaborative filtering approach in recommending advertising videos to both existing users and new users. From these results, we conclude that facial expressions are a critical factor for advertising video recommendations and are helpful in properly addressing the new user problem in existing recommender systems.

Research on Multidimensional Image Intelligent Matching Algorithm of Artificial Intelligence Image Recognition Technology

Image matching is a method of matching by analyzing the gray scale and texture information of the reference image and the image to be matched. Firstly, the scale invariant feature transform (SIFT) algorithm has long descriptor time and poor real time, a nonlinear dimension reduction method (LLE) based on local linear embedding is proposed to preserve the nonlinear information in the original data space as much as possible, shorten the running time of the algorithm, and improve the matching accuracy. Second, aiming at the problem that the Euclidean distance takes a large amount of calculation in the matching process, Manhattan distance is proposed to calculate the similarity between the reference image and the image to be matched, so as to further reduce the algorithm time. Through the improved LLE-SIFT algorithm, experimental results show that the algorithm has a high matching rate and improves the matching speed.

Dense Robust 3D Reconstruction and Measurement for 3D Printing Process Based on Vision

The 3D printing process lacks real-time inspection, which is still an open-loop manufacturing process, and the molding accuracy is low. Based on the 3D reconstruction theory of machine vision, in order to meet the applicability requirements of 3D printing process detection, a matching fusion method is proposed. The fast nearest neighbor (FNN) method is used to search matching point pairs. The matching point information of FFT-SIFT algorithm based on fast Fourier transform is superimposed with the matching point information of AKAZE algorithm, and then fused to obtain more dense feature point matching information and rich edge feature information. Combining incremental SFM algorithm with global SFM algorithm, an integrated SFM sparse point cloud reconstruction method is developed. The dense point cloud is reconstructed by PMVs algorithm, the point cloud model is meshed by Delaunay triangulation, and then the accurate 3D reconstruction model is obtained by texture mapping. The experimental results show that compared with the classical SIFT algorithm, the speed of feature extraction is increased by 25.0%, the number of feature matching is increased by 72%, and the relative error of 3D reconstruction results is about 0.014%, which is close to the theoretical error.

Feature Point Extraction and Motion Tracking of Cardiac Color Ultrasound under Improved Lucas–Kanade Algorithm

The purpose of this research is to study the application effect of Lucas–Kanade algorithm in right ventricular color Doppler ultrasound feature point extraction and motion tracking under the condition of scale invariant feature transform (SIFT). This study took the right ventricle as an example to analyze the extraction effect and calculation rate of SIFT algorithm and improved Lucas–Kanade algorithm. It was found that the calculation time before and after noise removal by the SIFT algorithm was 0.49 s and 0.46 s, respectively, and the number of extracted feature points was 703 and 698, respectively. The number of feature points extracted by the SIFT algorithm and the calculation time were significantly better than those of other algorithms ( P < 0.01 ). The mean logarithm of the matching points of the SIFT algorithm for order matching and reverse order matching was 20.54 and 20.46, respectively. The calculation time and the number of feature points for the SIFT speckle tracking method were 1198.85 s and 81, respectively, and those of the optical flow method were 3274.19 s and 80, respectively. The calculation time of the SIFT speckle tracking method was significantly lower than that of the optical flow method ( P < 0.05 ), and there was no statistical difference in the number of feature points between the SIFT speckle tracking method and the optical flow method ( P > 0.05 ). In conclusion, the improved Lucas–Kanade algorithm based on SIFT significantly improves the accuracy of feature extraction and motion tracking of color Doppler ultrasound, which shows the value of the algorithm in the clinical application of color Doppler ultrasound.