Related Experiment Video
Updated: Jul 7, 2026
08:25
Combining Eye-tracking Data with an Analysis of Video Content from Free-viewing a Video of a Walk in an Urban Park Environment
Published on: May 7, 2019
Feedback control strategies for object recognition
M Mirmehdi1, P L Palmer, J Kittler
1Sch. of Electr. Eng., Inf. Theor. and Math., Surrey Univ., Guildford, UK. M.Mirmehdi@cs.bris.ac.uk
Summary
This study introduces a novel feedback strategy for object detection, improving traditional methods. The approach enhances feature extraction and search efficiency, enabling accurate identification of objects in noisy conditions.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Image Processing
Background:
- Traditional image processing relies on single-pass hypothesis generation and verification.
- Classical systems face challenges with noisy data and complex object recognition.
Purpose of the Study:
- To present a new paradigm for feedback strategies in object detection.
- To improve upon established single-pass hypothesis generation and verification methods.
- To enable accurate object identification in challenging, noisy environments.
Main Methods:
- Implementing control strategies at low, intermediate, and high levels of analysis.
- Generating optimal low-level features to minimize hypotheses.
- Utilizing an interest operator for efficient hypothesis search and reduced false alarms.
- Employing feedback for updated feature extraction in noisy data.
Main Results:
- Successfully located target objects even in very noisy data.
- Minimized feedback to false alarms through optimal search direction.
- Achieved scale and rotation independent extraction of partially occluded objects.
- Demonstrated effectiveness using box-shaped objects in noisy infrared bridge images.
Conclusions:
- The proposed feedback strategy significantly enhances object detection capabilities.
- The system accurately identifies complex objects, even with partial occlusion and noise.
- This method offers a robust approach for object recognition in challenging image data.
Related Concept Videos
01:26Feedback control systems
Feedback control systems are categorized in various ways based on their design, analysis, and signal types.
Linear feedback systems are theoretical models that simplify analysis and design. These systems operate under the principle that their output is directly proportional to their input within certain ranges. For instance, an amplifier in a control system behaves linearly as long as the input signal remains within a specific range. However, most physical systems exhibit inherent nonlinearity...
Linear feedback systems are theoretical models that simplify analysis and design. These systems operate under the principle that their output is directly proportional to their input within certain ranges. For instance, an amplifier in a control system behaves linearly as long as the input signal remains within a specific range. However, most physical systems exhibit inherent nonlinearity...
01:10Control Systems
Control systems are everywhere in contemporary society, influencing diverse applications from aerospace to automated manufacturing. These systems can be found naturally within biological processes, such as blood sugar regulation and heart rate adjustment in response to stress, as well as in man-made systems like elevators and automated vehicles. A control system is essentially a network of subsystems and processes that collaboratively convert specific inputs into desired outputs.
At the heart...
At the heart...