ENHANCEMENT OF INFRARED IMAGE: A REVIEW

ANGAITKAR P.G.¹*, SAXENA K.^2
1Department Of Information Technology (M-Tech), Technocrats Institute of Technology, Bhopal, MP, India.
²Department Of Information Technology (M-Tech), Technocrats Institute of Technology, Bhopal, MP, India.
* Corresponding Author : pratikangaitkar70@gmail.com

Received : 21-02-2012     Accepted : 15-03-2012     Published : 19-03-2012
Volume : 3     Issue : 1       Pages : 76 - 78
J Signal Image Process 3.1 (2012):76-78

This paper provides an up-to-date critical survey of enhancement of infrared images by two popular algorithms namely spatial and spatiotemporal homomorphic filtering (SHF and STHF) based upon a infrared imaging model. Although spatiotemporal homomorphic filtering may reduce the number of iterations greatly in comparison to spatial one for a similar degree of convergence by making explicit use of the additional information provided temporally, the enhanced results from SHF are in general better than those from STHF. To provide this comprehensive survey we have not only focused on concept of infrared image but also two enhancement techniques are presented and discussed along with their advantages and disadvantages .

Infrared Images, Image Enhancement, Additive Wavelet Transform, Homomorphic Enhancement, Spatiotemporal homomorphic filtering.

Infrared (IR) light is electromagnetic radiation with a wavelength longer than that of visible light, measured from the nominal edge of visible red light at 0.74 micrometers, and extending conventionally to 300 micrometers. These wavelengths correspond to a frequency range of approximately 1 to 400 THz, [1] and include most of the thermal radiation emitted by objects near room temperature. Microscopically, IR light is typically emitted or absorbed by molecules when they change their rotational-vibration movements. Images received through various infrared (IR) devices in many applications are distorted due to the atmospheric aberration mainly because of atmospheric variations and aerosol turbulence [1,2] .

The International Commission on Illumination (CIE) recommended the division of infrared radiation into the following three bands:
• IR-A: 700 nm–1400 nm (0.7 µm – 1.4 µm, 215 THz - 430 THz).
• IR-B: 1400 nm–3000 nm (1.4 µm – 3 µm, 100 THz - 215 THz).
• IR-C: 3000 nm–1 mm (3 µm – 1000 µm, 300 GHz - 100 THz).

• A commonly used sub-division scheme is- Near-infrared (NIR, IR-A DIN). 0.75-1.4 µm in wavelength, defined by the water absorption, and commonly used in fiber optic telecommunication because of low attenuation losses in the SiO2 glass (silica) medium.
• Short-wavelength infrared (SWIR, IR-B DIN)- 1.4-3 µm in wavelength, water absorption increases significantly at 1,450 nm. The 1,530 to 1,560 nm range is the dominant spectral region for long-distance telecommunications.
• Mid-wavelength infrared (MWIR, IR-C DIN)- Also called intermediate infrared (IIR): 3-8 µm in wavelength. In guided missile technology the 3-5 µm portion of this band is the atmospheric window in which the homing heads of passive IR 'heat seeking' missiles are designed to work.
• Long-wavelength infrared (LWIR, IR-C DIN)- 8-15µm in wavelength. This is the "thermal imaging" region, in which sensors can obtain a completely passive picture of the outside world based on thermal emissions only and requiring no external light or thermal source such as the sun, moon or infrared illuminator.
• Far infrared (FIR)- SWIR is sometimes called "reflected infrared" while MWIR and LWIR is sometimes referred to as "thermal infrared." Due to the nature of the blackbody radiation curves, typical 'hot' objects, such as exhaust pipes, often appear brighter in the MW compared to the same object viewed in the LW.
The additive wavelet transform is the technique for enhancing the image and getting more information due to applying this technique. First decomposes an image into subbands using the “a’ trous” filtering approach [6–8] in several consecutive stages. The low pass filter used in this process has the following mask for all stages [8] .
Each difference between filter outputs of two consecutive stages is a subband of the original image. An image can be used represented as a product of tow components as in the following equation [8] :

f (n1, n2) = i (n1, n2 ) r (n1, n2) (1)

Where, f(n1, n2) is the obtained image pixel value, i(n1, n2) is the light illumination incident on the object to be imaged and r(n1, n2) is the reflectance of that object. It is known that illumination is approximately constant since the light falling on all objects is approximately the same. The only change between object images is in the reflectance component. If we apply a logarithmic process on Eq. (1), we can change the multiplication process into an addition process as follows:

log(f (n1, n2)) = log (i ((n1, n2)) + log (r (n1,n2)) (2)

The first term in the above equation has small variations but the second term has large variations as it corresponds to the reflectivity of the object to image. By attenuating the first the higher frequency subbands. The homomorphic enhancement algorithm uses for transforming these details to illumination and reflectance components. Then the reflectance components will be amplified showing the detail and clearly. Finally a wavelet reconstruction process is performed to get more informative, enhanced infrared image based on characteristics such as scale, shape, texture, pixel value dynamic range, pixel value statistical distribution, spatial frequency, brightness, and contrast differences.
The enhancement of infrared images is slightly different from traditional image enhancement in dealing with the large black areas and the small details. So suggested approach aims at separating the details in different subbands and processing each subband, separately. We have found that the additive wavelet transform is a powerful tool in image decomposition. If the infrared image is decomposed using the additive wavelet transform, the details can be separated into the higher frequency subbands. Due use the homomorphic enhancement algorithm for transforming these details to illumination and reflectance components.
The reflectance components are amplified showing the details and clearly. Finally a wavelet reconstruction process is performed to get an enhanced infrared image with much more details.

The two algorithms called spatial and spatiotemporal homomorphic Filtering i.e. SHF and STHF were designed. The second one makes explicit use of the temporal information provided by the image sequence so that an enhancement faster than that obtained by utilizing the spatial homomorphic Filtering was obtained. STHF will spend much lower iterations and much fewer time to obtain a resulting images for a similar degree of convergences the enhance images are in general not as good as those from SHF.
Dim target detection is particularly challenging because standard techniques such as spatial, thresholding, and edge detection can fail due to the lack of contrast between target and background. To detect targets in clutter, there must be a set of characteristics that can be exploited to separate targets from clutter. Separation can be based on characteristics such as scale, shape, texture, pixel value dynamic range, pixel value statistical distribution, spatial frequency, brightness, and contrast differences.
Qi, H. and J. F. Head [1] suggested that Wavelets have demonstrated some effectiveness for target detection. Traditionally, there are four primary applications of wavelet-based methods to target detection: These are as Follows:
• Wavelets as edge detectors.
• Using wavelets to separate targets from clutter based on scale differences.
• Using wavelets as approximate matched filters.
• Capturing target dynamic range differences using wavelet filters.
Using wavelets as edge detectors assumes that target edges differ in some way from clutter edges. For example, edges from natural clutter may be more diffuse whereas edges from man-made objects such as vehicles may be harder, sharper, and more distinct. Kuruganti, P. T and H. Qi [2] and Scales, N., C. Herry, and M. Frize [3] stated as edge detectors, wavelets may be designed to capture these edge differences. Zhang, C. J., F. Yang [4] stated that targets may also differ from clutter by characteristic scalesAndreone, L., P. C. Antonello and M. Bertozzi [5] stated that wavelets can be also designed to function as approximate matched filters. For such usage, wavelet filters are designed to produce a large response when matched against a target region. Lastly, the low and high pass filters from the wavelet decomposition can be used to detect target regions of low or high pixel value dynamic range.

A. Spatial homomorphic Filtering (SHF)
• IT Requires more iteration i.e. more time
• Enhance image quality is not much good.
B. Spatiotemporal homomorphic Filtering (STHF)
• IT requires lower iteration i.e. less time as compare to SHF.
• Enhance Image quality is much good as compare to SHF.
C. Additive Wavelet Transform
• Very few Iteration as compare to SHF & STHF.
• This technique is very fast as compare to SHF & STHF.

• Wave length of Infrared Image is Low.
• Infrared enabled devices are not cost efficient

• Night Vision Images- Infrared is used in night vision equipment when there is insufficient visible light to see. Night vision devices operate through a process involving the conversion of ambient light photons into electrons which are then amplified by a chemical and electrical process and then converted back into visible light.
• Heating- Infrared heating is also becoming more popular in industrial manufacturing processes, e.g. curing of coatings, forming of plastics, annealing, plastic welding, print drying. In these applications, infrared heaters replace convection ovens and contact heating.
• Meteorology- Weather satellites equipped with scanning radiometers produce thermal or infrared images which can then enable a trained analyst to determine cloud heights and types, to calculate land and surface water temperatures, and to locate ocean surface features.

This review paper presents new techniques to enhancing the infrared image i.e. SHF, SHTF and additive wavelet Transform, the homomorphic enhancement. The homomorphic processing is applied to the infrared image subbands separately. Then, these subbands are merged again to reconstruct an enhanced image [8] .

[1] Qi H. and Head J.F. (2002) Second Joint EMBS/BMES Conference.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[2] Kuruganti P.T. and Qi H. (2002) Second Joint EMBS/BMES Conference.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[3] Scales N., Herry C. and Frize M. (2004) 26th Annual International Conference of the IEEE EMBS.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[4] Zhang C.J., Yang F., Wang X.D. and Zhang H.R. (2005) Fourth International Conference on Machine Learning and Cybernetics.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[5] Andreone L., Antonello P.C., Bertozzi M., Broggi A., Fascioli A. and Ranzato D. (2002) The IEEE 5th International Conference on Intelligent Transportation Systems.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[6] Zhang C., Wang X., Zhang H., Lv G. and Wei H. (2006) First International Conference on Innovative Computing, Information and Control.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[7] Uvais Qidwai (2008) IEEE Trans on Image Processing, 17(8).  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

[8] Ashiba H.I., Awadallah K.H., El-Halfawy S.M. and Abd El-Samie F.E. (2008) Progress In Electromagnetic Research C, 1, 123-131.  
» CrossRef   » Google Scholar   » PubMed   » DOAJ   » CAS   » Scopus  

	Fig. 1- Additive wavelet decomposition
	Fig. 2- Wavelet reconstruction