Image Constraint Technique Used by Bio-Inspired TSA Optimized

Algorithm for Large Memory Management

Sachin More

and Dayanand Ingle

Bharati Vidyapeeth College of Engineering, Mumbai University, CBD Belapur, Navi Mumbai, India

Keywords: Image quality, Image Compression, Resolution Scalable Image Compression, Quality Scalable Image

Compression, Highly Scalable Image Compression and TSA Optimized Discrete Wavelet Transform.

Abstract: The well-known classical image compression algorithms are renowned for their outstanding performance,

low computational complexity, and production of an embedded bitstream that can be decoded at different bit

rates with progressively better picture quality more bits are received. It takes a lot of computer memory since

it stores the created sets and the pixel locations in linked lists. The transform's multiresolution capability does

not lead to the generation of a resolution-scalable bitstream that would enable different resolutions to decode

the image. The TSA Optimized Discrete Wavelet Transform approach, which only needs one list of fixed size

measuring 1/4 the size of the image and an average of 2.25 bits/pixel, solves the massive memory problem of

traditional techniques. The study's main contribution is to scale up image quality and resolution in order to

free up more memory. The TSA Optimized Discrete Wavelet Transform approach is recommended in this

study to decrease complexity and increase speed. Due to the algorithm's special characteristics, it is ideal for

addressing the different skills and resolution related preferences of today’s Internet users.

1 INTRODUCTION

The fundamental purpose of lossy image compression

is to decrease the distinction between the unique and

recovered photographs even as also reducing the

common quantity of bits in step with the pixel of the

compressed photo as a good deal as viable. The MSE

among those pictures, or suggest squared errors, is

usually used to quantify the disparity. The

computational complexity and reminiscence use of an

image compression method are extra important

components. For this reason, most of these elements

need to be considered for any algorithm to be truly

judged

(Jayasankar, 2012).

A picture may be recovered the use of a traditional

photo compression scheme at a single bitrate

(pleasant) and resolution (size). This compression

paradigm might not work for all customers because

the existing purchaser base has a wide range of

bandwidth, show resolution, computing power, and

memory capacities. Moreover, clients want to rapid

look through all the searched photographs while

viewing snapshots at the net in advance than choosing

https://orcid.org/0000-0003-4077-2251

https://orcid.org/0000-0002-2654-3279

the specified image(s). Adopting this approach of

picture compression will bring about time,

bandwidth, memory, and processing electricity

losses. But a scalable image compression gadget

permits the setup person to decode the photo at an

appropriate excellent or/and determination through

controlling the recovered image's outstanding or/and

resolution (length). Most effective the first-rate. By

the usage of the usage of some sorts a QSIC may be

completed. However, the selection of the recovered

image may be modified (Taubman, 2002). Due to the

HSIC's potential to decode snap shots of various best

and resolutions, it's miles specifically appealing to the

desires of cutting-edge clients.

A picture may be recovered the use of a traditional

image compression scheme at an single bitrate (nice)

and backbone (length). This compression paradigm

may not art work for all customers due to the reality

the prevailing character base has a great type of

bandwidth, show selection, computing strength, and

reminiscence capacities. Furthermore, consumers

want to fast look through all the searched snap shots

More, S. and Ingle, D.

Image Constraint Technique Used by Bio-Inspired TSA Optimized Algor ithm for Large Memory Management.

DOI: 10.5220/0013606900004664

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 3rd International Conference on Futuristic Technology (INCOFT 2025) - Volume 2, pages 897-903

ISBN: 978-989-758-763-4

897

Figure 1: Conventional Image Compression Diagram

while viewing pics on the net earlier than selecting

the required image(s). Certainly, adopting this

technique of photo compression will bring about

time, bandwidth, memory, and processing strength

losses. Instead, a scalable image compression tool

allows the end person to decode the photo at the

suitable extraordinary or/and determination by way of

manner of controlling the recovered photo

satisfactory or/and determination (duration) (Said and

Pearlman, 1996; Lee and Hung, 2019). Because of the

HSIC's capability to decode photographs of diverse,

first-rate resolutions, it is mainly attractive for the

needs of contemporary customers.

The aim of this commentary was modified to

signify a picture compression technique that would

permit large reminiscence space to be maintained

while lowering reminiscence length. The endorsed

compression approach consists of an single image

compression approach and a bit-feet approach for

keeping the compression ratio and uniformity of

photograph excellence at some point of the photo.

The suggested machine, at the facet of the bit fee

management technique, continues widely talking

using the facts collected inside the contemporary

frame. So that you can similarly improve picture

excellence, the approach, also determines series. This

may be completed by using something as

fundamental as the photograph. It can drastically

reduce the problem of image-satisfying imbalance

photograph, a flaw in low-reminiscence compression

strategies that is unavoidable.

The remaining sections of the paper are as

follows: Background material and common

compression techniques are given in Section 2 of this

essay. Section 3 provides a description of the

suggested compression scheme with memory

management. The approach is assessed and the

findings are discussed in Section 4. This essay

concludes in Section 5.

2 LITERATURE SURVEY

Based on zero tree capabilities of wavelet

coefficients, the 1D SPIHT approach (Lee and Hung,

2019) is used. This technique creates 3 dynamic lists

via frequently crossing better bit to the decrease bit

plane to send important bit facts first, enhancing

compression efficiency. This method is simple to use

and proof in opposition to transmission faults.

Because of its prolonged memory accesses at some

point of the repetitious manner coding, the 1D SPIHT

compression technique, nevertheless, incurs good

sized energy consumption

Figure 2: Different Techniques of Image Compression.

The visually lossless compression approach of

line compression emerges as located out in

(Umbaugh, 2023). Whilst errors are harder to

identify, this approach might also carry out median

component detection (MED) prediction of 2 4 blocks.

It could also perform MED prediction of one–four

blocks in complexity and diploma places in which

errors are apparent. After the information has been in

addition compressed through way of a preset non-

uniform quantization step. A frequency adaptive line

compression (FALC) method with less complexity in

evaluation to modern compression techniques has

been suggested in (Chew, 2009). The FALC approach

uses entire after four stages of wavelet adjustments.

The fourth step involves selective zero-area

quantization within the four immoderate-frequency

bands. Redundancy inside the low-frequency

variability has eliminated using predictive coding.

Then, using a Huffman coding-based totally variable

duration coding (VLC) technique, data from every

frequency band are compressed. With a visually

lossless nation, FALC's compression ratio exceeds

3:1. To actively keep more power in contexts in

which IoT programs are used, but, extra advancement

is needed.

INCOFT 2025 - International Conference on Futuristic Technology

898

Because of the reality every line, even inside a

single frame, has precise homes in phrases of

compression effectiveness, the picture high-quality of

the road-based totally compression algorithms

regularly modifications at some stage in a single

body. Both Jiang's and Edirisinghe's line-based

totally bit charge manipulates (BRC) were proposed

for use in the JPEG-LS context (Deepthi and Prasad,

2018). But, in terms of uniformity inside a single

frame, those BRCs lack ok photograph well.

Primarily based at the cutting-edge-day FALC

method, a one-of-a-type bit rate management

mechanism (Al-Janabi and Harbi, 2022) has been

cautioned to decorate the consistency of photograph

fantastic inside one frame. The quantization degree

shift is saved sluggish. Whilst thirds of the frame have

been processed, it additionally evaluates the picture

exquisite and the closing bits to decide the proper

quantization degree to be able to use up the very last

bits.

However, modern-day BRC strategies

nonetheless warfare with inconsistent image

satisfactory, especially whilst scenes or pics are

separated. Consequently, TSA desires to be more

suitable greater earlier than it may be used for

consumer goods

3 PROPOSED METHODOLOGY

The TSA Optimized Discrete Wavelet remodel

techniques are to begin with brought on this section.

In comparison to the alternative techniques, the TSA

Optimized Discrete Wavelet redecorate, set of

guidelines is drastically much less complicated and

plays slightly higher. The number one contribution of

the research is then provided, that is the advised TSA

Optimized Discrete Wavelet rework method for

memory control. A particularly scalable bitstream,

this is each high-quality and backbone scalable in

addition to less in duration and taking on greater

RAM is created by using the usage of the TSA

Optimized Discrete Wavelet remodel set of

regulations.

One of the population-based algorithms is TSA

(Tree Seed Algorithm), and the population will be

referred to as stand moving forward. Eq. 1 (Meraj and

Khan, 2021) is used to derive the stand.

a =𝑇

,

=𝐿



+𝑟

,

×𝑈



−𝐿



, 𝑖 =

1,2,…….,𝑁 𝑎𝑛𝑑

𝑗

=1,2,…..,𝐷

)

Figure 3: Proposed Block Diagram of this Work.

Following the production of the bushes within the

stand, the fitness values of the trees are decided the

usage of a goal characteristic tailor-made to the

optimization mission.

Via utilizing Eq. 2 or three, a positive range of

seeds is generated for every tree at the same time as a

preset termination circumstance is glad (Al-Janabi,

2013)

𝑆

,

=𝑇

,

+𝛼

,

×(𝐵



−𝑇

,

) (2)

𝑆

,

=𝑇

,

+𝛼

,

×(𝑇

,

−𝑇

,

) (3)

The choice of Eqs. 2 or 3 relies upon on a TSA

manage parameter. The ST (search tendency)

manipulate parameter has a variety of [0,1]. If it is

miles vital to intensify the stand across the ideal tree

placement, a high fee for the ST parameter should be

selected. If the stand's exploration ability is extended,

the ST parameter should be set to a low fee. A random

wide variety within the range [0,1] is created for the

selection of equations 2 or 3. The Eq. 2 is selected if

this fee is much less than ST; in any other case, the

Eq. Three is chosen. In a nutshell, the ST parameter

is utilized to modify the stand's potential for

exploration and exploitation. The usage of the

number of seeds can in addition beautify the local

seek abilities around the tree websites. The minimum

and the greatest number of seeds in the TSA

algorithm are 10% of the timber and 25% of the trees,

respectively. Therefore, based totally at the stand, the

number of seeds is determined (Alam and Khan,

2012).

Image Constraint Technique Used by Bio-Inspired TSA Optimized Algorithm for Large Memory Management

899

The quest method and equations defined above

are iterated sequentially after the set of rules is

initialized, if a predefined termination situation is

happy.

• Step 1 select the general populace length

(filter out window length), ST Values, and

termination situation.

• Step 2: Initialize the populace in accordance

with the supplied Eqs. (2) and (3).

• Step 3: the usage of the Eqs. (3) based on the

values of ri;j and ST, create seed values for

the modern tree values. If the seed values

supplied progressed health in comparison to

the unique tree values, replace the tree values.

• Step 4: the usage of Eq. (3), pick out the

premier cost and confirm the termination

circumstance.

• Step 5: Assign the optimized values and

record the outcome if the termination

situation is happy; in any other case, repeat

steps 2, three, and four

Figure 4: Tree Seed Algorithm Flowchart.

1. Select the best axolotls, according to

the

unction

2. For each axolotl with optimization

value 1≤

𝑗

≤

𝑀

Compute inverse probability of

transition

If each component I of the current

axolotl as n

Update the optimization value

Update

𝑓



Even as a tree generates new seed web sites, there

can be several seeds present; the variety of seeds

depends on the dimensions of the population. The

bottom wide type of seeds that may be generated

through a set of rules inside the right manner. Eq. (3)

is utilized for if a randomly generated integer in the

range [0, 1] is smaller than ST

The suggested method for selecting the best

compression, placing is depicted in determining

Preceding to sending records to the receiver, the

endorsed approach can shop for separately

compressed image statistics using the further

effective compression method is selected because the

output value thru the comparator in the middle, which

forecasts and compares the bits produced through

each compression mode. For images with extra stated

characteristic changes among picture traces, together

with the recommended technique can make sure a

more compression overall performance.

The counselled solution tries to apply get a higher

compression ratio; it then chooses this sort of

predictive coding options that is minimal in length.

Due to the fact differential values exhibited much less

correlations inside the vertical course after making

use of the horizontal predictive coding, we did not use

each prediction modes simultaneously.

Based at the average strength of the H1 bands

from the preceding body, the cautioned approach also

forecasts an regular compression outline for the

present frame. Two centre subsections of the previous

frame's many subsections are examined to look if a

split mode exists. Break up mode may be dealt with

by means of way of the recommended TSA optimized

algorithm technique in a single body. In among

successive frames, it could additionally deal with

scene modifications. If resource of an additional than

a threshold cost (TH) in advance than the closing

quarter of the frame, the scene has altered. If a scene

change happens interior a body, the default TCR

charge (4.00) is used to compress the closing areas.

Moreover, through manner of the usage of break up

mode, this method helps forestall unneeded image

INCOFT 2025 - International Conference on Futuristic Technology

900

best deterioration delivered on by scene changes

among frames.

4 RESULTS AND DISCUSSION

Because of the several problematic interactions

between applications, this section can estimate the

electricity intake of an set of rules running on a goal

embedded system whilst warding off considerable

estimation deviation (Calderbank, 1998). The mean

squared blunders (MSE) among the preliminary N

pixels is used to gauge overall performance. Image

(Io) and the reconstructed image (Ir), each of size M,

are described as MSE as (Hu and Yang, 2020):

𝑀𝑆𝐸 =

𝑀𝑁

[𝐼



(

𝑖,

𝑗

)

−𝐼



(𝑖,

𝑗

)

]











)

𝑃𝑆𝑁𝑅 = 10𝑙𝑜𝑔











𝑑𝑒𝑐𝑖𝑏𝑒𝑙

(

𝑑𝑏

)

(5)

𝑀𝐸𝑀



== 2𝑏

(

2𝑁



)

+2𝑐











=4𝑏𝑁



𝑐













4𝑏 +







𝑁



𝑏𝑖𝑡 (6)

Table 1: PSNR values evaluated for existing and proposed

algorithm.

Bit

rate

(

Bpp)

Encoding time

(

seconds

)

Decoding time

(

seconds

)

Lena Goldhill

SPIHT

(Harbi,2

022)

SLS

(Sara,

2019)

TSA

(Prop

osed)

SPIHT

(Harbi,

2022)

SLS

(Sara,

2019)

TSA

(Propo

sed)

0.06 0.74 0.72 0.52 0.34 0.068 0.020

0.11 1.17 1.14 1.40 0.561 0.082 0.051

0.15 1.06 1.05 1.00 1.000 0.077 0.130

0.5 2.21 2.20 2.15 1.717 0.112 0.175

1 4.60 4.55 3.45 3.400 0.445 0.237

Table 2: Memory necessities of TSA, LHS-TSA, and the

proposed HS-MSLS for different image sizes (Wu and

Wang, 2013)

Image

size

Memory (kb)

TSA LHS-TSA HS-MSLS

(Song, 2017)

MEM

(kb)

% MEM

(kb)

% MEM

(kb)

256×256

273 2.12 31 0.24 45 0.35

512×512

1270 2.40 127 0.24 200 0.38

1024×1024

5685 2.67 511 0.24 653 0.41

Table 3: Current and planned algorithms, complexity is expressed by the encoding time and decoding time against the bit

rate

Bit

rate

(Bpp)

PSNR (db)

Lena Goldhill Barbara

SPIHT

(Harbi,

2022)

SLS

(Sara,

2019)

TSA

(Proposed)

SPIHT

(Harbi,

2022)

SLS

(Sara,

2019)

TSA

(Proposed)

SPIHT

(Harbi,

2022)

SLS

(Sara,

2019)

TSA

(Proposed)

0.0615 25.61 26.46 27.69 25.16 25.18 25.25 22.17 22.20 22.50

0.115 28.38 30.14 31.19 26.66 26.69 26.75 23.26 23.31 23.46

0.15 31.61 32.00 34.17 28.65 28.67 28.78 25.86 25.89 25.93

0.5 35.12 35.20 37.41 31.16 31.18 31.30 30.87 30.90 30.96

1 38.40 38.51 39.72 34.41 34.45 34.56 35.15 35.21 35.35

Image Constraint Technique Used by Bio-Inspired TSA Optimized Algorithm for Large Memory Management

901

5 CONCLUSIONS

In this research, a compression approach with bit rate

control was presented for size reduction in memory

management contexts. Given the power constraints of

the memory management environment, the suggested

TSA algorithm employs a low-complexity approach

with the fewest number of processes and memory

accesses. Based on the results of our tests, the

suggested solution outperformed existing methods in

terms of compression ratio while keeping a lower

level of complexity. In terms of size reduction effect

at the system level, it performed better than existing

SPIHT or 4L HEVC methods. In cases where there is

a high/low quality imbalance, picture split mode, and

scene transitions, which frequently occur in

conventional line compression methods, it may retain

superior restorative and constant image quality

performance.

REFERENCES

Jayasankar, U., Thirumal, V., & Ponnurangam, D. (2021).

A survey on data compression techniques: From the

perspective of data quality, coding schemes, data type

and applications. Journal of King Saud University-

Computer and Information Sciences, 33(2), 119-140/

Thanki, R. M., & Kothari, A. (2019). Hybrid and advanced

compression techniques for medical images. Springer

International Publishing

Taubman, D., Ordentlich, E., Weinberger, M., & Seroussi,

G. (2002). Embedded block coding in JPEG

2000. Signal Processing: Image

Communication, 17(1), 49-72.

Rüefenacht, D., Naman, A. T., Mathew, R., & Taubman, D.

(2019). Base-anchored model for highly scalable and

accessible compression of Multiview imagery. IEEE

Transactions on Image Processing, 28(7), 3205-3218.

Van Fleet, P. J. (2019). Daubechies Wavelet

Transformations.

Vetterli, M. (1995). Wavelets and sub band coding.

Said, A., & Pearlman, W. A. (1996). A new, fast, and

efficient image codec based on set partitioning in

hierarchical trees. IEEE Transactions on circuits and

systems for video technology, 6(3), 243-250.

Lee, R. C., & Hung, K. C. (2019). New modified SPIHT

algorithm for data compression system. Journal of

Medical and Biological Engineering, 39, 18-26.

Khan, E., & Ghanbari, M. (2004). An efficient and scalable

low bit-rate video coding with virtual SPIHT. Signal

Processing: Image Communication, 19(3), 267-283.

Senapati, R. K., Pati, U. C., & Mahapatra, K. K. (2012).

Listless block-tree set partitioning algorithm for very

low bit rate embedded image compression. AEU-

International Journal of Electronics and

Communications, 66(12), 985-995.

Umbaugh, S. E. (2023). Digital image processing and

analysis: computer vision and image analysis. CRC

Press.

Chew, L. W., Ang, L. M., & Seng, K. P. (2009, August).

Reduced memory spiht coding using wavelet transform

with post-processing. In 2009 International Conference

on Intelligent Human-Machine Systems and

Cybernetics (Vol. 1, pp. 371-374). IEEE.

Deepthi, S. A., Rao, E. S., & Prasad, M. G. (2018, January).

Image transmission and compression techniques using

SPIHT and EZW in WSN. In 2018 2nd International

Conference on Inventive Systems and Control

(ICISC) (pp. 1146-1149). IEEE.

Alam, M., Khan, E., & Gopal, B. (2012). Modified listless

set partitioning in hierarchical trees (MLS) for memory

constrained image coding applications. Current Trends

in Signal Processing, 2(2), 56-66.

Al-Janabi, A. K., Al-Musawi, H. K., & Harbi, Y. J. (2022).

An efficient and highly scalable listless SPIHT image

compression framework. Journal of applied research

and technology, 20(2), 173-187.

Meraj, Y., & Khan, E. (2021, March). Modified ZM-

SPECK: A low complexity and low memory wavelet

image coder for VS/IoT Nodes. In 2021 International

Conference on Emerging Smart Computing and

Informatics (ESCI) (pp. 494-500). IEEE.

Al-Janabi, A. K. (2013). Low memory set-partitioning in

hierarchical trees image compression

algorithm. International Journal of Video & Image

Processing and Network Security IJVIPNS-

IJENS, 13(2), 12-18.

Alam, M., & Khan, E. (2012). Listless Highly scalable set

partitioning in hierarchical trees coding for

transmission of image over heterogenous

networks. International Journal of Computer

Networking, Wireless Mobile Commun.

Danyali, H., & Mertins, A. (2004). Flexible, highly

scalable, object-based wavelet image compression

algorithm for network applications. IEE Proceedings-

Vision, Image and Signal Processing, 151(6), 498-510.

Calderbank, A. R., Daubechies, I., Sweldens, W., & Yeo,

B. L. (1998). Wavelet transforms that map integers to

integers. Applied and computational harmonic

analysis, 5(3), 332-369.

Hu, Y., Yang, S., Yang, W., Duan, L. Y., & Liu, J. (2020,

July). Towards coding for human and machine vision:

A scalable image coding approach. In 2020 IEEE

International Conference on Multimedia and Expo

(ICME) (pp. 1-6). IEEE.

Al-Janabi, A. K., Al-Musawi, H. K., & Harbi, Y. J. (2022).

An efficient and highly scalable listless SPIHT image

compression framework. Journal of applied research

and technology, 20(2), 173-187.

Sara, U., Akter, M., & Uddin, M. S. (2019). Image quality

assessment through FSIM, SSIM, MSE and PSNR—a

comparative study. Journal of Computer and

Communications, 7(3), 8-18.

Wu, D., Zhang, H., Li, X., & Wang, J. (2013, June).

Multiview Video Coding Based on Wavelet Pyramids.

INCOFT 2025 - International Conference on Futuristic Technology

902

In 2013 International Conference on Computational

and Information Sciences (pp. 225-228). IEEE.

Song, X., Huang, Q., Chang, S., He, J., & Wang, H. (2017).

Three‐dimensional separate descendant‐based SPIHT

algorithm for fast compression of high‐resolution

medical image sequences. IET Image

Processing, 11(1), 80-87.

Image Constraint Technique Used by Bio-Inspired TSA Optimized Algorithm for Large Memory Management

903