Modular Anti-noise Deep Learning Network for Robotic Grasp Detection Based on RGB Images

Zhaocong Li

2023-10-30

Abstract:While traditional methods relies on depth sensors, the current trend leans towards utilizing cost-effective RGB images, despite their absence of depth cues. This paper introduces an interesting approach to detect grasping pose from a single RGB image. To this end, we propose a modular learning network augmented with grasp detection and semantic segmentation, tailored for robots equipped with parallel-plate grippers. Our network not only identifies graspable objects but also fuses prior grasp analyses with semantic segmentation, thereby boosting grasp detection precision. Significantly, our design exhibits resilience, adeptly handling blurred and noisy visuals. Key contributions encompass a trainable network for grasp detection from RGB images, a modular design facilitating feasible grasp implementation, and an architecture robust against common image distortions. We demonstrate the feasibility and accuracy of our proposed approach through practical experiments and evaluations.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The paper attempts to address the problem of accurately detecting the pose of graspable objects from a single RGB image in robotic grasping tasks. Traditional methods often rely on depth sensors to obtain depth information, but these sensors are costly and have limited usage scenarios. With the advancement of technology, more and more research has begun to shift towards using more affordable and accessible RGB images for grasp detection. However, RGB images lack depth information and are susceptible to noise, blur, and other image quality issues in practical applications, posing new challenges for grasp detection.

To tackle these challenges, the paper proposes a method based on a modular noise-resistant deep learning network. This method not only detects graspable objects from a single RGB image but also improves the accuracy of grasp detection by integrating semantic segmentation information. Additionally, the network design is highly robust, maintaining high performance when processing blurred and noisy images. The main contributions of the paper include:

1. Proposing a trainable neural network that can detect graspable objects from a single RGB image.
2. Designing a modular network structure that integrates multiple training components to achieve feasible grasp detection implementation.
3. Constructing a robust architecture capable of handling blurred and noisy images.

Experimental validation shows that this method performs excellently in practical applications, maintaining high accuracy and robustness even when dealing with noisy images.