College of Engineering, Pune

About: College of Engineering, Pune is a based out in . It is known for research contribution in the topics: Sliding mode control & Control theory. The organization has 4264 authors who have published 3492 publications receiving 19371 citations. The organization is also known as: COEP.

A novel and cost effective resistive rain sensor for automatic wiper control: Circuit modelling and implementation

Abstract: Safety and driver comfort are the essential goals of new trends in automobile industry. An automatic wiper controller, helps not only in increasing safety by reducing distractions but also increases the overall comfort. Such an automatic control is available however it has limitations of high cost and low efficiency. In this work, we have proposed an automatic wiper controller, based on resistive rain sensor which is cost effective, efficient and has a wide range of output. An equivalent electrical and mathematical model of the sensor is developed, simulated and practically verified. The rain sensor has a predetermined geometry. Hence the rain water forms a film on the sensor surface causing its resistance to change non-linearly. To increase the overall efficiency of system, it is necessary to linearize the sensor response. It is achieved using electrical equivalent model of the sensor and with appropriate linearization circuit. Further, a customized embedded system is developed using PIC micro-controller to achieve motor speed variation based on sensor output. The proposed system is rigorously tested on a small segment car (Indica V2) for its performance. It is observed that, the developed system works satisfactorily under various rain scenarios.

Dynamic hand gesture recognition using kinect sensor

Abstract: Continuous trajectory path made by hand over a period of time is considered for gesture recognition purpose. Dynamic gesture recognition has always been critical, since conventional method utilize separation of hand from surrounding environment and then finding of palm points make it hard. In this paper real time dynamic hand gesture recognition system using kinect sensor has been proposed. In this the movement of human hand will be captured by kinect sensor and it will be processed in OpenCV. Signal generated by OpenCV are given to Arduino and it will be utilized for controlling DC servo motor action. The computer will generate appropriate signals for controlling the action of robotic arm. The movements of robotic arm links are controlled by DC servo motors. Real-time hand tracking technique is used for object detection in the range of vision.

Integer & fractional order PID Controller for fractional order subsystems of AUV

Abstract: Control and modeling of nonlinear systems like Autonomous Underwater Vehicle (AUV) is a challenge for researchers. Here the modeling and controller scheme designed is based on fractional order system. In this article Fractional order proportional integral derivative (FOPID) controller is designed and applied to fractional order depth and steering systems of AUV and it is compared with classical integer order PID Controller.

E-voting system using visual cryptography & secure multi-party computation

Abstract: The paper aims at creation of secure voting system using visual cryptography and secure multi-party computation. Visual Cryptography is the method of encrypting the visual data for authentication of the voter. Visual cryptography scheme which is used to create 2 shares of fingerprint image. This system prevents voters for voting which try to vote using fake voter identification card (VIC). This system is developed robustly to ensure that all eligible voters having a voter identification card (VIC) is allowed to cast their respective vote. One share is stored at administrator side and one share is stored in VIC, so no one can access the full fingerprint image. For security of votes the principle of secure multi-party computation is used. Secure multi-party computation allows multiple parties to participate in a computation. Security, accuracy reliability and transparency are the major concern in these systems. The voters, who cast multiple votes during the voting process is ensured to be prevented by biometric identification of the votes could be used for casting their vote and restricting them to cast again. The process of Electronic-voting system could be deployed with four phases - the voter's enrollment, voter's authentication, vote casting and recording, and vote counting and election result publication.

Segmentation using region growing algorithm based on CLAHE for medical images

Abstract: As medical images are mostly fuzzy in nature, segmentation of intensity based image is the most challenging task. A novel region growing segmentation algorithm for the image segmentation to detect tumor is presented where selective median filter is used for pre-processing. For enhancement Contrast Limited Adaptive Histogram Equalization (CLAHE) method is used. It is found that in the images processed with CLAHE, lesions appear obvious to the background and the image detail is very good. It is useful for radiologists to see subtle edge information, such as speculation. Next Scale Invariant Feature transform is used to automatically find growing seeds and the seeded region growing rule for the development of regions is able to detect both strong and weak edges, which is more suitable compared to other methods.