An Accurate Multi-Row Panorama Generation Using Multi-Point Joint Stitching

BASE PAPER TITLE: An Accurate Multi-Row Panorama Generation Using Multi-Point Joint Stitching

OUR PROPOSED CONCEPT: Panorama Generation using SURF

ABSTRACT:

Most of the existing panorama generation tools require the input images to be captured along one direction, and yield a narrow strip panorama. To generate a large viewing field panorama, this paper proposes a multi-row panorama generation (MRPG) method. For a pan/tilt camera whose scanning path covers a wide range of horizontal and vertical views, the image frames in different views correspond to different coordinate benchmarks and different projections. And the image frame should not only be aligned with the continuous frames in timeline but also be aligned with other frames in spatial neighborhood even with long time intervals. For these problems, MRPG first designs an optimal scanning path to cover the large viewing field, and chooses the center frame as the reference frame to start to stitch. The stitching order of multi-frame is arranged in first-column and second-row to ensure a small alignment error. Moreover, MRPG proposes a multi-point joint stitching method to eliminate the seams and correct the distortions, which makes the current frame accurately integrated into the panoramic canvas from all directions. Experimental results show that MRPG can generate a more accurate panorama than other state-of-the-art image stitching methods, and give a better visual effect for a large viewing field panorama.

PROJECT OUTPUT VIDEO:

EXISTING SYSTEM:

• There are many panoramic image stitching technologies in use nowadays. For example, we can easily use our mobile phones or digital cameras to generate a panoramic photo, or we can use some commercial applications, such as Autostitch, Kolor Autopano, Microsoft ICE, Realviz3 and Microsoft Photosynth, to synthesize several images and generate a panorama. These experiences give us an illusion that image stitching is a mature technology.

DISADVANTAGES OF EXISTING SYSTEM:

• Although panoramic image stitching technologies have been widely used, the existing image stitching algorithms still have a lot of deficiencies. The most obvious problem is the limited coverage of the viewing field.
• The misalignment results in the image blur and dislocation.

PROPOSED SYSTEM:

• In this paper, the work focuses on an accurate multi-row panorama generation method, which uses multiple images to generate a wide viewing field panorama. For the given start point and terminal point, this paper firstly proposes an optimal scanning path to cover the whole viewing field between the two points. According to the path, the center frame is selected as the reference frame, and the coordinate of the reference frame is used as the benchmark to produce the coordinate of the panorama. It helps to avoid strabismus and reduce cumulative errors. And then, the stitching order is arranged in first-column and second-row to ensure a small alignment error.
• Moreover, a multi-block and multi-point joint stitching method is proposed. Multi-block refers to calculate the projective transformation model for each image block to solve the problem that the different blocks in a large size panorama have the different projections.

ADVANTAGES OF PROPOSED SYSTEM:

• Multi-point refers to neighbor constrained points and relative constrained points, and these points help to align multi-frame from all directions. Neighbor constrained points are use to eliminate the cutting seams existing in neighboring blocks, and relative constrained points are used to maintain the correct relative positions between the blocks, which make the wide viewing panorama look coincident and accurate.

SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS: 

• System : Pentium Dual Core.
• Hard Disk : 120 GB.
• Monitor : 15’’ LED
• Input Devices : Keyboard, Mouse
• Ram :2 GB

SOFTWARE REQUIREMENTS: 

• Operating system : Windows 10
• Coding Language :
• Tool : MATLAB 2019

REFERENCE:

JIN ZHENG (Member, IEEE), ZHI ZHANG, QIUHAO TAO, KAI SHEN, AND YUE WANG, “An Accurate Multi-Row Panorama Generation Using Multi-Point Joint Stitching”, IEEE ACCESS, 2018.