Research on the Extraction of Image Edge Information in Convolutional Neural Networks

Convolutional neural network is an important neural network model in deep learning and a common algorithm in computer vision problems. From the perspective of practical application scenarios, this paper studies whether padding in convolutional neural network convolution layer weakens the image edge information. In order to eliminate the background factor, this paper select MNIST dataset as the research object, move the 0-9 digital image to the specified image edge by clearing the white area pixels in the specified direction, and use OpenCV to realize bilinear interpolation to scale the image to ensure that the image dimension is 28×28. The convolution neural network is built to train the original dataset and the processed dataset, and the accuracy rates are 0.9892 and 0.1082 respectively. In the comparative experiment, padding cannot solve the problem of weakening the image edge weight well. In the actual digital recognition scene, it is necessary to consider whether the core recognition area in the input image is at the edge of the image.

Moving computational multi-domain method for modelling the flow interaction of multiple moving objects

This study proposes a method where the flow field variables are communicated between multiple separate moving computational domains and simulates the flow interaction of multiple moving objects. Instead of using the conventional approach with a single fixed computational domain covering the whole flow field, this method advances the moving computational domain (MCD) method in which the computational domain itself moves in line with the motions of an object inside. The computational domains created around each object move independently, and the flow fields of each domain interact where the flows cross. This eliminates the spatial restriction for simulating multiple moving objects. After the results of the shock tube test verify that the interpolation has been achieved between grids, a validation test is conducted in which two spheres are crossed, and the forces exerted on one object due to the other’s crossing at a short distance are calculated. The results verify the reliability of this method and show that it is applicable to the flow interaction of multiple moving objects.

Human–Robot Collaborative Assembly Based on Eye-Hand and a Finite State Machine in a Virtual Environment

With the development of the global economy, the demand for manufacturing is increasing. Accordingly, human–robot collaborative assembly has become a research hotspot. This paper aims to solve the efficiency problems inherent in traditional human-machine collaboration. Based on eye–hand and finite state machines, a collaborative assembly method is proposed. The method determines the human’s intention by collecting posture and eye data, which can control a robot to grasp an object, move it, and perform co-assembly. The robot’s automatic path planning is based on a probabilistic roadmap planner. Virtual reality tests show that the proposed method is more efficient than traditional methods.

Effect of Grouping, Segmentation, and Vestibular Stimulation on the Autokinetic Effect

We report some new observations on what could be regarded as the world’s simplest visual illusion—the autokinetic effect. When a single dim spot of light is viewed in a completely dark room, it moves vividly in random directions. During steady fixation, perhaps subtle eye movements cause the image to move and a failure to correct for this using eye movement command signals leads to motion perception. This is especially true because eye muscle fatigue can lead to miscalibration. However, if two dots are shown, they often move independently in different directions, which negate the eye movement theory. In addition, two lines defining a single cross sometimes uncouple and slide past each other and the fragments composing a hidden object move independently until they click in place and the whole object is perceived—implying that the illusion occurs relatively late in visual processing. Finally, the effect is modulated by vestibular stimulation; anchoring your sense of self may be a prerequisite for binding features into coherent objects.