Vrushali Rajaram Pagire

Bio: Vrushali Rajaram Pagire is an academic researcher from Massachusetts Institute of Technology. The author has contributed to research in topics: Object detection & Trajectory. The author has an hindex of 2, co-authored 4 publications receiving 7 citations. Previous affiliations of Vrushali Rajaram Pagire include College of Engineering, Pune.

FPGA based moving object detection

Abstract: The detection of moving object is fundamental to automated video surveillance, object analysis and recognition. It involves the trajectory of an object over time by locating its position in every frame of the video. In this paper we implement a moving object detection method on FPGA. Detection of moving object is a challenging task. In our implementation of the design we used the background subtraction based method. Background subtraction is a method which detects moving objects from reference background image. Experiment is carried out in MATLAB and on FPGA board.

Detection of Moving Object Based on FPGA

Abstract: Detecting a moving object is very crucial in surveillance application. Many algorithms have been developed for object detection, recognition and analysis. The algorithms involve the trajectory of detected object by locating position in each captured image. In the proposed work a system is implemented to detect foreground object in any environmental condition and the model is implemented on FPGA board. Real time foreground object detection is a very complex task. Proposed system detects object by background subtraction algorithm based on dynamic thresholding. Background subtraction algorithm detects moving object by subtracting current frame from reference background frame. The system is simulated using MATLAB and implemented on FPGA board.

Underwater Moving Object Detection Using GMG

Abstract: The development in ocean exploration and observation make the demand for moving object detection in underwater increasingly urgent. The moving object detection in underwater with dynamic or moving background is the challenging and difficult task for the researchers. In dynamic environment, foreground object as well as background of the scene both are in motion condition, thus it is very difficult to detect foreground moving object in underwater medium. In proposed work, background is categorized into three types i.e. static, moderate dynamic and high dynamic backgrounds. The GMG algorithm invented by Andrew B. Godbehere, Akihiro Matsukawa, Ken Goldberg is implemented to detect moving object and compare the results for the three background categories. The sensitivity and accuracy parameter values are considered for the comparison of the results.

Underwater Fish Detection and Classification using Deep Learning

Abstract: The researchers face a difficult problem in detecting and identifying underwater fish species. Marine researchers and ecologists must evaluate the comparative profusion of fish species in their environments on a regular basis and track population trends. Researchers have presented a number of underwater computer vision, machine learning-based automatic systems for fish detection and classification. However, due of the changing undersea environment, it is extremely challenging to find the ideal system for detecting and classifying fish. Because light has such a strong influence in the aqueous medium, conducting research in this environment is difficult. The MobileNet model is utilised to detect and recognise the fish breed in the proposed work. The dataset is preprocessed before the model is implemented in order to obtain appropriate performance metrics. The work is based on the Kaggle dataset, which has nine different fish breeds in total. With a 99.74 percent accuracy, the model can detect and recognise nine different breeds. In comparison to other state of art methods, the model exhibits promising results.

Navigation of Underwater Remotely Operated Vehicle

Abstract: Remotely operated vehicles (ROVs) are underwaterrobots that are commanded from the surface by a person. This robot is made up of embedded systems and mechanical systems, and it is connected to the outside world via a series ofload-bearing umbilical cables that hold power, data, and communication cables. For the ROV, data transfer as well ascontrol signals are required. The navigation team was made up of working on the ROV's navigation system. It has a video camera and a ballast system for balance, movement propulsion system pumps, and IMU to locate the position of a vehicle microcontrollers to compute, process, and offer the navigation system, sensor for feedback data back to the lighting and components controlling and coordinating all of this is a challenge. A manipulator, subsea equipment, and devices to assess clarity, temperature, and depth are all included in the proposed task. The purpose of the proposed effort is to create such a ROV to execute a given mission involving several persons. In addition, a novel form of ROV attitude control is introduced, which makes use of floats to modify the ROV's centre of buoyancy. This is an interdisciplinary endeavour in which the members are from many fields. Together with the mechanical department, the electronics navigation team worked on this.

Hardware implementation of a background substraction algorithm in FPGA-based platforms

Abstract: Different strategies for the implementation of a fuzzy logic-based background subtraction algorithm are presented in this paper. The goal of this contribution is to obtain an efficient implementation suitable to be integrated into hardware platforms with limited resources. In order to find an adequate performance-resources trade-off, the design space is explored taken into account several strategies and implementation options. The final implementation is encapsulated within an IP core that has been used in a demonstrator, built on a Spartan-3A-DSP FPGA development board, suitable for processing VGA (640×480P) @ 60 Hz.

Underwater Moving Object Detection Using GMG

Abstract: The development in ocean exploration and observation make the demand for moving object detection in underwater increasingly urgent. The moving object detection in underwater with dynamic or moving background is the challenging and difficult task for the researchers. In dynamic environment, foreground object as well as background of the scene both are in motion condition, thus it is very difficult to detect foreground moving object in underwater medium. In proposed work, background is categorized into three types i.e. static, moderate dynamic and high dynamic backgrounds. The GMG algorithm invented by Andrew B. Godbehere, Akihiro Matsukawa, Ken Goldberg is implemented to detect moving object and compare the results for the three background categories. The sensitivity and accuracy parameter values are considered for the comparison of the results.

Real-Time Moving Target Detection System Based on FPGA

Abstract: With the rapid development of social economy, video surveillance technology has attracted more and more attention. The proliferation of video data has made the calculation and storage of image data more difficult. However, to process images with software cannot meet the growing demand of video business because it is slow. Current moving target detection algorithms consume a lot of resources and power in PC (Personal Computer). In view of these problems, this paper aims to deploy the motion target detection algorithm on the ZYNQ hardware platform with the combination of hardware and software. On the ZYNQ hardware platform, the video image acquisition, frame difference algorithm, morphological image processing and bounding box design are completed in the PL (Programmable Logic) part, and the camera and VDMA (Video Direct Memory Access) related configuration are completed in the PS (Processing System) part. The experimental results show that the system can detect moving targets in real time with lower power consumption and fewer resources, which can be applied to the field of video surveillance.

Automated Crop Field Surveillance using Computer Vision

Abstract: Artificial Intelligence is everywhere today. But unfortunately, Agriculture has not been able to get that much attention from Artificial Intelligence (AI). A lack of automation persists in the agriculture industry. For over many years, farmers and crop field owners have been facing a problem of trespassing of wild animals for which no feasible solution has been provided. Installing a fence or barrier like structure is neither feasible nor efficient due to the large areas covered by the fields. Also, if the landowner can afford to build a wall or barrier, government policies for building walls are often very irksome. The paper intends to give a simple intelligible solution to the problem with Automated Crop Field Surveillance using Computer Vision. The solution will significantly reduce the cost of crops destroyed annually and completely automate the security of the field.

Underwater Fish Detection and Classification using Deep Learning

Abstract: The researchers face a difficult problem in detecting and identifying underwater fish species. Marine researchers and ecologists must evaluate the comparative profusion of fish species in their environments on a regular basis and track population trends. Researchers have presented a number of underwater computer vision, machine learning-based automatic systems for fish detection and classification. However, due of the changing undersea environment, it is extremely challenging to find the ideal system for detecting and classifying fish. Because light has such a strong influence in the aqueous medium, conducting research in this environment is difficult. The MobileNet model is utilised to detect and recognise the fish breed in the proposed work. The dataset is preprocessed before the model is implemented in order to obtain appropriate performance metrics. The work is based on the Kaggle dataset, which has nine different fish breeds in total. With a 99.74 percent accuracy, the model can detect and recognise nine different breeds. In comparison to other state of art methods, the model exhibits promising results.