Dhruvil Darji

95% accuracy was achieved to detect the road cracks. Convolutional Neural Network(CNN) approach has used to get this accuracy. CNN has better accuracy than conventional technics of Image processing.
Data segmentation and Localization have been preprocessed. Used Tensorflow Environment. Writing a paper on it.
Same Idea has applied to find the phone in the images. 90.7% accuracy was achieved to localize the phone, which is enough to find a thief.

32 bit single cycle microprocessor using VHDL code . Used Xilinx software for simulation. Processor could perform million instructions per second using 5 stage pipeline system.

Noisy signal was given, by doing convolution process using filter noise can be removed. So, I wrote code in visual studio and simulate it using Matlab.

It's real time project. Clock of 1 sec was implemented using vhdl. Implemented different conditions of traffic at intersection of cross roads and traffic light behave according to traffic. I made controller to control different possibilities of traffic and traffic light operates on basis of that. Simulation was done by Xilinx Software.

Using booth's algorithm implemented multiplier using different logic gates in vhdl language. I also made controller so it work with all the possible 32 bit integer numbers. Xilinx was used for simulation.

93% accuracy has been achieved to count the right number of targets. Communication between multiple drones has established with optimum power usage.
Targets appeared more than once in the images has been optimizes by target patterns in different images.
Autonomous flights of the drones have been planned successfully. The flight is based on GPS.
I got FAA rule-based UAV Remote Pilot 107 certification.
Next target is to optimize the flight paths looking at density of the targets with… Show more

93% accuracy has been achieved to count the right number of targets. Communication between multiple drones has established with optimum power usage.
Targets appeared more than once in the images has been optimizes by target patterns in different images.
Autonomous flights of the drones have been planned successfully. The flight is based on GPS.
I got FAA rule-based UAV Remote Pilot 107 certification.
Next target is to optimize the flight paths looking at density of the targets with each drones. Show less

It is the robot, which can follow path itself and reach destination. it works 8085 microprocessor with object sensors.

it works on gprs + gsm devices with ARM processor based on lpc2148 controller. I interfaced 3 sensors with it. it could sense object , fire and third one is switch. when any of the sensor will detect, it will switch gprs and gsm active and through that will get text on seleted devices, as well as police and fire station will also get notified with it.