High-resolution dynamic visual observation method and device
A high-resolution, high-resolution technology, applied in neural learning methods, biological neural network models, neural architectures, etc., can solve the problem of low spatial resolution of event cameras and achieve the effect of improving efficiency
Active Publication Date: 2020-09-22
TSINGHUA UNIV
View PDF5 Cites 6 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0006] In order to solve the problem of low spatial resolution of existing event cameras, the present invention provides a high-resolution dynamic visual observation method and device, which can obtain higher Spatial high-resolution event data
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0023] The following combination image 3 The technical scheme of the invention is described in further detail.
[0024] Such as image 3 As shown, this embodiment provides a high-resolution dynamic visual observation method for event stream super-resolution and scene high-resolution observation, including the following steps:
[0025] Step 1. Obtain low-resolution event stream data, and obtain low-spatial-resolution event stream data from the event camera;
[0026] In this step, the existing event camera is used to collect low spatial resolution event flow data to be super-resolutioned. As a new type of bionic camera, the event camera outputs data when and only when the light intensity of the scene changes. This output data is called an event stream. The event stream data output by the event camera can be expressed in the form of formula (1)
[0027]
[0028] Among them, ε is the set of event flow, i represents the number of the event in the event flow, x i is the spa...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a high-resolution dynamic visual observation method which comprises the following steps: 1, acquiring low-resolution event stream data, and acquiring low-spatial-resolution event stream data from an event camera; 2, performing data preprocessing, obtaining the spatial resolution size of the high-resolution event data according to the spatial resolution size of the input low-resolution event data and expectation, and obtaining the multiple of the resolution needing to be improved through calculation; 3, constructing a spiking neural network, and constructing a full-connection spiking neural network by using a Spike Response model as a neuronal dynamics model; 4, completing the calculation of a high-resolution space-time pulse signal; and 5, carrying out data post-processing to realize high-resolution dynamic visual observation. According to the method, event data with higher spatial resolution can be obtained from sparse event data with low resolution, and the effect of high-resolution dynamic visual observation is achieved.
Description
technical field [0001] The invention relates to the fields of computer vision and neuromorphic computing, in particular to a high-resolution dynamic visual observation method and device. Background technique [0002] An event camera is a bio-inspired sensor that works very differently than a traditional camera. Unlike traditional cameras that collect the absolute light intensity of a scene at a fixed frame rate, this camera outputs data if and only when the scene light intensity changes. This output data is called an event stream. Compared with traditional cameras, event cameras have the advantages of high dynamic range, high temporal resolution, and no motion blur. [0003] As a new type of visual sensor, the event camera has a completely different output data form from traditional cameras, and cannot directly apply various algorithms of traditional cameras and images. Traditional cameras collect the light intensity value of the scene at a fixed rate (that is, frame rate)...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06N3/04G06N3/08G06N3/063
CPCG06N3/049G06N3/08G06N3/063
Inventor 高跃李思奇
Owner TSINGHUA UNIV