scispace - formally typeset
Search or ask a question
Topic

Human visual system model

About: Human visual system model is a research topic. Over the lifetime, 8697 publications have been published within this topic receiving 259440 citations.


Papers
More filters
Proceedings ArticleDOI
01 Dec 2013
TL;DR: A new color constancy model is introduced by imitating the functional properties of the HVS from the retina to the double-opponent cells in V1 with competitive results in comparison to the complex state-of-the-art approaches, but with a simple implementation and without the need for training.
Abstract: The double-opponent color-sensitive cells in the primary visual cortex (V1) of the human visual system (HVS) have long been recognized as the physiological basis of color constancy. We introduce a new color constancy model by imitating the functional properties of the HVS from the retina to the double-opponent cells in V1. The idea behind the model originates from the observation that the color distribution of the responses of double-opponent cells to the input color-biased images coincides well with the light source direction. Then the true illuminant color of a scene is easily estimated by searching for the maxima of the separate RGB channels of the responses of double-opponent cells in the RGB space. Our systematical experimental evaluations on two commonly used image datasets show that the proposed model can produce competitive results in comparison to the complex state-of-the-art approaches, but with a simple implementation and without the need for training.

45 citations

Patent
18 Feb 2000
TL;DR: In this paper, a method for lossy compression of images reduces visual distortion for a given compressed bitrate or, equivalently, requires a lower bit-rate for the given level of visual distortion, where the metric is sensitive to masking properties of the human visual system.
Abstract: A method for lossy compression of images reduces visual distortion for a given compressed bit-rate or, equivalently, requires a lower bit-rate for a given level of visual distortion. An image is decomposed using a space-frequency transform and frequency bands are then partitioned into small blocks. The blocks are independently quantized and coded using an embedded block coder, so that each block bit-stream contains a large number of finely spaced truncation points. A visual distortion measure is computed for each block at each truncation point, where the metric is sensitive to masking properties of the Human Visual System. The distortion values and bit-stream lengths corresponding to each block's truncation point are used to optimise overall visual distortion at one or more target bit-rates or to minimise the bit-rate corresponding to one or more target visual distortion levels. A computationally and memory efficient procedure is described for computing the visual distortion measure for each block's truncation point, within each frequency band, as required by the subject compression system.

45 citations

Journal ArticleDOI
TL;DR: A generic and flexible system for vision-based robot control that integrates visual tracking and visual servoing approaches in a unifying framework using a template matching algorithm based on an efficient second-order minimization.

45 citations

Proceedings ArticleDOI
04 Oct 1994
TL;DR: It is argued that a single visual mechanism called "zooming" addresses these scaling problems and, when suitably augmented, can also support automatic component discovery and intelligent error correction.
Abstract: Visual programming research has largely focused on the issues of visual programming-in-the-small. However, entirely different concerns arise when one is programming-in-the-large. We present a visual software engineering environment that allows users to construct visually programs consisting of hierarchically organized networks of components that process streams of arbitrary objects. We discuss the problems that occur when trying to construct systems consisting of thousands of interconnected components, examine how this environment deals with some of the problems specific to visual programming-in-the-large, and show why our initial solutions failed to scale successfully. Finally, we argue that a single visual mechanism called "zooming" addresses these scaling problems and, when suitably augmented, can also support automatic component discovery and intelligent error correction. >

45 citations

Proceedings ArticleDOI
01 Oct 2013
TL;DR: A new FIve-Step BLInd Metric (FISBLIM) is proposed, built upon several common image processing blocks to simulate the image perceiving process of the human visual system (HVS).
Abstract: The last decade has seen a surge of interest in the research of image quality assessment (IQA). Many successful quality metrics, such as structural similarity index (SSIM) are reportedly to achieve very high accuracy for various kinds of image distortions. However, in practice, multiple image distortions tend to occur together and this leads difficulty to previous works of IQA including SSIM and variations. This problem is even more difficult for no-reference or blind quality assessment. To answer this challenge, this paper proposes a new FIve-Step BLInd Metric (FISBLIM) for quality assessment of multiply distorted images. The algorithm is built upon several common image processing blocks to simulate the image perceiving process of the human visual system (HVS). The FISBLIM method is not training based and the performance is robust and not database-dependent. Experimental results on the newly released LIVE multiply distorted image quality database demonstrate the effectiveness of FISBLIM as compared with mainstream full-reference and no-reference image quality metrics.

45 citations


Network Information
Related Topics (5)
Feature (computer vision)
128.2K papers, 1.7M citations
89% related
Feature extraction
111.8K papers, 2.1M citations
86% related
Image segmentation
79.6K papers, 1.8M citations
86% related
Image processing
229.9K papers, 3.5M citations
85% related
Convolutional neural network
74.7K papers, 2M citations
84% related
Performance
Metrics
No. of papers in the topic in previous years
YearPapers
202349
202294
2021279
2020311
2019351
2018348