Microplastics (MPs, size ≤ 5 mm) have emerged as a significant worldwide concern, threatening marine and freshwater ecosystems, and the lack of MP detection technologies is notable. The main goal of this research is the development of a camera sensor for the detection of MPs and measuring their size and velocity while in motion. This study introduces a novel methodology involving computer vision and artificial intelligence (AI) for the detection of MPs. Three different camera systems, including fixed-focus 2D and autofocus (2D and 3D), were implemented and compared. A YOLOv5-based object detection model was used to detect MPs in the captured image. DeepSORT was then implemented for tracking MPs through consecutive images. In real-time testing in a laboratory flume setting, the precision in MP counting was found to be 97%, and during field testing in a local river, the precision was 96%. This study provides foundational insights into utilizing AI for detecting MPs in different environmental settings, contributing to more effective efforts and strategies for managing and mitigating MP pollution.

Microplastics (<5 mm) have become a global concern due to their growing threat to the marine and freshwater environment. There is a lack of technologies for the rapid and accurate identification and quantification of microplastics in the aqueous environment. This paper presents a deep-learning-based methodology for real-time detection, tracking, and counting of microplastics in freshwater environments through real-time object detection. A prototype was developed to detect microplastics of 1 mm to 5 mm in size and different shapes (e.g., spherical) and colors (e.g., red, green, blue). The microplastics detection model employed the small YOLOv5 architecture as we focused on low-power applications. In-situ image collection was performed using a Logitech C270 camera, and the microplastics were manually annotated on those images before being applied for model training. For real-time object tracking, we used Simple Online and Real-time Tracking with a Deep Association Metric (DeepSORT), an extended version of the Simple Online and Real-time Tracking (SORT) algorithm. Our developed system can work up to 34 cm/sec of water velocity and successfully detect, track, count, and calculate the velocity of microplastic of size 5mm.

This paper presents a smart wheelchair, a customized version of a commercially available manual one. The motivation of this research was to support individuals with limited mobility in navigating their world. The proposed automated wheelchair consists of a stereo depth camera, LIDAR, DC gear motor, motor controller, joystick, emergency switch, processors, etc. The users have the option of controlling where their wheelchair goes by navigating themselves through the joystick or automatically through the developed hands-free application. This app relays back to the Jetson processor situated underneath the chair, which has the capability to move through any indoor location to reach the user's chosen destination. Although the initial research goal was to enable the wheelchair to navigate a patient through an indoor hospital facility, this low-cost, lightweight wheelchair is envisioned to be employed in all indoor locations and accessible by a broader range of users than just the wealthy few. The project's design files have been made available so that this open-source wheelchair can be reproduced publicly.

A wheelchair can provide limited but crucial mobility to an injured or disabled individual. This paper presents the first stage of the development of a smart wheelchair which is the customization of a manually controlled wheelchair with a novel implementation of octascopic vision. This relatively inexpensive design of an autonomous wheelchair consists of two monochromic camera arrays (each having four cameras) placed around the frame of the wheelchair to achieve a view of 360 degrees. The initial research goal was to design a wheelchair controlled by the embedded processor, allowing the wheelchair to navigate autonomously around an indoor facility with and without human intervention. Additionally, it was intended to allow those previously denied access to the world of automatic wheelchairs because of a low personal income. Through the testing of wheelchair functionality, (a) a large dataset of octascopic images was captured from this wheelchair, and (b) a YOLOv7-based object detection model was developed to avoid obstacles and autonomously control the movement. This paper presents the camera placement and the obstacle detection model using octascopic images. All the project design files have been granted an open-source license and can be reproduced publicly.

This paper presents the development of a Simultaneous Localization and Mapping (SLAM) based Autonomous Navigation system. The motivation for this study was to find a solution for navigating interior spaces autonomously. Interior navigation is challenging as it can be forever evolving. Solving this issue is necessary for a multitude of services, like cleaning, the health industry, and in manufacturing industries. The focus of this paper is the description of the SLAM-based software architecture developed for this proposed autonomous system. A potential application of this system, oriented to a smart wheelchair, was evaluated. Current interior navigation solutions require some sort of guiding line, like a black line on the floor. With this proposed solution, interiors do not require renovation to accommodate this solution. The source code of this application has been made open source so that it could be re-purposed for a similar application. Also, this open-source project is envisioned to be improved by the broad open-source community upon past its current state.

Implementing vision-based automated control in sophisticated robotics (i.e., prosthetic hands) is limited due to their high-power consumption and computational extensiveness, which may not be achievable through employing commercially available microcontrollers and their coprocessors. Hence, a digital design from scratch and the involvement of the most recent power-efficient but advanced FPGA processors are required. This paper presents the first stage of such an FPGA implementation of an advanced image processing pipeline where an RGB camera is interfaced to a Zynq® UltraScale+™ MPSoC, where an object detection deep learning model is hosted. Acknowledging the complexity of this pipeline, a thorough description of the implementation is provided with a shared public repository. Optimizing the design and circuitry will ensure the implementation of vision sensors to artificially sense an object from the wrist position and facilitate sophisticated grabbing by a prosthetic hand.