In this paper, we show a system that can automatically track the human body and recognize its motion by making the camera move using Pan-Tilt HAT. Skeleton information is extracted from images through Human Pose Estimation technology and the relationship is identified through it. And the range of recognition can be increased by designating the motion of the camera so that the human body is at the center.
Moon Classification based on 2 level Hough Circle Transform
Since the shape of the moon is sufficiently characterized, it can be classified using image processing-based technology. In this paper, we show a method of detecting the area of the moon by dividing the Hough circle transform into two stages and classifying it using the ratio to the area. Experimental results demonstrate that the canny performs better than adaptive thresholding when the hough circle transform is used twice and in the preprocessing process to find the region.
Typhoon Route Prediction using Bidirectional Recurrent NeuralNetwork
2020. 11
Hyeoncheol Son, Daseul Kim, Moonnyeon Kim and Sungyoung Kim
This paper introduces a bidirectional recurrent neural network that predicts the future route of the typhoon. The recurrent neural networks are affected by the typhoon’s previous route and predict future route. By designing this recurrent neural network in both directions, it provides higher-accuracy.
Converting Close-Looped Electronic Circuit Image with Single I/O Symbol into Netlist
In this paper, we have developed a new method to convert drawing images of electronic circuits into PSPICE Netlists.
First, circuit symbols are recognized through learning about circuit symbols.
We then use image processing to analyze the relationship between wires connecting the recognized circuit symbols based on the drawing where the symbols were removed. Finally, we construct Netlists by combining the symbol recognition and wire analysis information.
The results of the proposed method show high performance on cyclic electronic circuit diagrams consisting of a single I/O.
Wire Removal and Recognition on Circuit Elements on Electronic Schematics
While new electronic schematics may need to be created, existing
electronic schematics are often reused. It is a very tedious process to redraw the
schematic by hand to recreate the drawings that exist in printed form on paper.
It would be very convenient if existing drawings drawn on paper could be
automatically recognized and digitized without user intervention. In this paper,
we present a method of detecting wires and recognizing several important
circuit elements from existing electronic circuit diagrams. This paper is a
preliminary study to analyze existing electronic schematics and convert them
into PSpice's script code.