09-07-2011, 12:19 PM
[attachment=14382]
ABSTRACT
The objective of this project is to develop the matlab implementation for performing the following functions:
• To acquire the displacement and velocity information of moving vehicles such as military tanks
• To identify the type of target with in the range from 100m to 3km
• To estimate the movements of the vehicle from tracking system
The sequence of steps during development are as follows:
• Two consecutive frames of source image which is captured online using CCD is taken. The basic unit is developed for 64X64 2-D image segments. This can be generalized for 512X512 images.
• The estimation of real time velocity information of the objects is performed by frame subtraction and X,Y histograms methods using MATLAB signal processing tool box.
• Modeling of this problem is developed in for the calculation of real time histogram of a digital image.
• The online captured image data frames are processed “frame by frame” .
• Using this technique, the velocity of moving targets can be estimated with the accuracy of ±1m.The results obtained from this project can be used in electronic warfare system.
CHAPTER-1
INTRODUCTION
1.1 Objectives of the project
This project deals with the development and implementation of system on chip for object tracking using histograms.
The objective of the project includes the following functions:
To acquire the distance and velocity information of moving vehicles such as military tanks.
To identify the type of target within the range from 100m to 3Km
To estimate the movements of the vehicle.
1.2 Overview of the project
This project deals with the development and implementation of image Histogrammer used for calculation for Infrared Images. These 2-D histograms are useful for position determination of desired objects in the images obtained from CCD Cameras. The problem is oriented towards the processing of monochrome images. Further, image segments of 64 x 64 pixels are considered for conceptual design. The same coding and implementation could be generalized to 512 x 512 pixel images.
The actual problem is defined and the solution methodology has been, initially, developed in MATLAB. This conceptual modeling enables us to define the problem clearly and quickly. Further, the time optimization of techniques used could be verified
Fig1.1: Block diagram of object tracking control system
The main application of Image Histogram is to Estimation the Velocity of moving objects, which are in the range of 100 meters to 3000 meters away from the tracking system. The technique is especially useful in tracking of military tanks in missile applications. The dynamic motion of the missile could be tracked and controlled through the feedback loop. The CPU control system is shown in fig. 1.1. In this application, the Image Histogrammer constitutes the Image Processing block.
1.3 Theoretical Description of Velocity Estimation using Image Processing
The principle of 2-D image histogram can be used to estimate the velocity of a moving object (military tanks). Velocity estimation plays an important role in automatic tracking of military tanks in defense applications. The identification of the target is based on Pattern Recognition.
One of the approaches used for velocity estimation and pattern recognition is by the analysis of the images of the moving targets by using image-processing techniques. Image processing involves processing and analyzing the images by various image-processing techniques depending upon the application. These images are generated by a CCD (Charge Coupled Device) camera.
Two consecutive image data frames are stored in the RAM processing. The processing involves pixel by pixel operation. The subtracted values are stored in the third frame. Then X, Y histograms are calculated for the resulting image. Then the moving average filtering (smoothing) is done over histograms. Finally to obtain the maxima “ Find maxima index” algorithm has been developed. The difference between two maxims of successive frames x-histogram gives the ∆x and the difference between two maxims of successive frames y-histogram gives the ∆y, which can be used to calculate the velocity of moving objects.
The difference in recording times of the two frames gives the time interval of the motion of object as ∆t = T2 – T1
Velocity of the objects in X-direction (Vx) = dx / (t2 – t1)
Velocity of the objects in Y-direction (Vy) = dy / (t2-t1)
Where t2-t1 = ∆t = 1 / (frame frequency).
where T is the time taken by the object to move from first position to second position, that can be calculated from the frame frequency (30 frames/sec), i.e.,
T = 1 / Frame frequency
∆x indicates the displacement moved horizontally with respect to the previous frame in ∆t time.
∆y indicates the displacement moved vertically with respect to the previous frame in ∆t time.
ABSTRACT
The objective of this project is to develop the matlab implementation for performing the following functions:
• To acquire the displacement and velocity information of moving vehicles such as military tanks
• To identify the type of target with in the range from 100m to 3km
• To estimate the movements of the vehicle from tracking system
The sequence of steps during development are as follows:
• Two consecutive frames of source image which is captured online using CCD is taken. The basic unit is developed for 64X64 2-D image segments. This can be generalized for 512X512 images.
• The estimation of real time velocity information of the objects is performed by frame subtraction and X,Y histograms methods using MATLAB signal processing tool box.
• Modeling of this problem is developed in for the calculation of real time histogram of a digital image.
• The online captured image data frames are processed “frame by frame” .
• Using this technique, the velocity of moving targets can be estimated with the accuracy of ±1m.The results obtained from this project can be used in electronic warfare system.
CHAPTER-1
INTRODUCTION
1.1 Objectives of the project
This project deals with the development and implementation of system on chip for object tracking using histograms.
The objective of the project includes the following functions:
To acquire the distance and velocity information of moving vehicles such as military tanks.
To identify the type of target within the range from 100m to 3Km
To estimate the movements of the vehicle.
1.2 Overview of the project
This project deals with the development and implementation of image Histogrammer used for calculation for Infrared Images. These 2-D histograms are useful for position determination of desired objects in the images obtained from CCD Cameras. The problem is oriented towards the processing of monochrome images. Further, image segments of 64 x 64 pixels are considered for conceptual design. The same coding and implementation could be generalized to 512 x 512 pixel images.
The actual problem is defined and the solution methodology has been, initially, developed in MATLAB. This conceptual modeling enables us to define the problem clearly and quickly. Further, the time optimization of techniques used could be verified
Fig1.1: Block diagram of object tracking control system
The main application of Image Histogram is to Estimation the Velocity of moving objects, which are in the range of 100 meters to 3000 meters away from the tracking system. The technique is especially useful in tracking of military tanks in missile applications. The dynamic motion of the missile could be tracked and controlled through the feedback loop. The CPU control system is shown in fig. 1.1. In this application, the Image Histogrammer constitutes the Image Processing block.
1.3 Theoretical Description of Velocity Estimation using Image Processing
The principle of 2-D image histogram can be used to estimate the velocity of a moving object (military tanks). Velocity estimation plays an important role in automatic tracking of military tanks in defense applications. The identification of the target is based on Pattern Recognition.
One of the approaches used for velocity estimation and pattern recognition is by the analysis of the images of the moving targets by using image-processing techniques. Image processing involves processing and analyzing the images by various image-processing techniques depending upon the application. These images are generated by a CCD (Charge Coupled Device) camera.
Two consecutive image data frames are stored in the RAM processing. The processing involves pixel by pixel operation. The subtracted values are stored in the third frame. Then X, Y histograms are calculated for the resulting image. Then the moving average filtering (smoothing) is done over histograms. Finally to obtain the maxima “ Find maxima index” algorithm has been developed. The difference between two maxims of successive frames x-histogram gives the ∆x and the difference between two maxims of successive frames y-histogram gives the ∆y, which can be used to calculate the velocity of moving objects.
The difference in recording times of the two frames gives the time interval of the motion of object as ∆t = T2 – T1
Velocity of the objects in X-direction (Vx) = dx / (t2 – t1)
Velocity of the objects in Y-direction (Vy) = dy / (t2-t1)
Where t2-t1 = ∆t = 1 / (frame frequency).
where T is the time taken by the object to move from first position to second position, that can be calculated from the frame frequency (30 frames/sec), i.e.,
T = 1 / Frame frequency
∆x indicates the displacement moved horizontally with respect to the previous frame in ∆t time.
∆y indicates the displacement moved vertically with respect to the previous frame in ∆t time.