Goal-directed evaluation of binarization methods

Trier, Øivind Due; Jain, Anil K.

doi:10.1109/34.476511

Cited by 530 publications

(212 citation statements)

References 26 publications

Supporting

Mentioning

208

Contrasting

Unclassified

Order By: Relevance

“…Some are based on the local variance of the image intensity. More specifically, the threshold for each pixel is calculated according to the local mean and variance in a window of a pre-determined size [25]- [27]. The formulation is simple and straight forward; however, the results are easily corrupted by the presence of spurious local intensity maxima.…”

Section: Introductionmentioning

confidence: 99%

A novel computational approach for automatic dendrite spines detection in two-photon laser scan microscopy

Cheng

Zhou

Miller

et al. 2007

Journal of Neuroscience Methods

View full text Add to dashboard Cite

Background-Recent research has shown that there is a strong correlation between the functional properties of a neuron and its morphologic structure. Current morphologic analyses typically involve a significant component of computer-assisted manual labor, which is very time-consuming and is susceptible to operator bias. The existing semi-automatic approaches largely reduce user efforts. However, some manual interventions, such as setting a global threshold for segmentation, are still needed during image processing.Methods-We present an automated approach, which can greatly help neurobiologists obtain quantitative morphological information about a neuron and its spines. The automation includes an adaptive thresholding method, which can yield better segment results than the prevalent global thresholding method. It also introduces an efficient backbone extraction method, a SNR-based, detached spine component detection method, and an attached spine component detection method based on the estimation of local dendrite morphology.Results-The morphology information obtained both manually and automatically are compared in detail. Using the Kolmogov-Smirnov test, we find a 99.13% probability that the dendrite length distributions are the same for the automatic and manual processing methods. The spine detection results are also compared with other existing semi-automatic approaches. The comparison results show that our approach has 33% fewer false positives and 77% fewer false negatives on average.Conclusions-Because the proposed detection algorithm requires less user input and performs better than existing algorithms, our approach can quickly and accurately process neuron images without user intervention.

show abstract

Section: Introductionmentioning

confidence: 99%

A novel computational approach for automatic dendrite spines detection in two-photon laser scan microscopy

Cheng

Zhou

Miller

et al. 2007

Journal of Neuroscience Methods

View full text Add to dashboard Cite

show abstract

“…To conserve local details and handle local illumination level one requires small window size but if choose too small window size then it will not cover object and eliminate noise in the gray image. Window size of 15 9 15 and k = -0.2 was recommended by Trier and Jain [5].…”

Section: Niblack's Methodsmentioning

confidence: 99%

“…Trier Jain [5] evaluated goal directed binarization Methods and shown that all well known existing algorithms work good on constant illuminated document images but fails in case of varying illumination. In general, global thresholding performs well for documents where there is a clear separation between foreground and background.…”

Section: Introductionmentioning

confidence: 99%

Efficient binarization technique for severely degraded document images

Singh

Mridula

2014

CSIT

View full text Add to dashboard Cite

Degradations in document images appear due to shadows, non-uniform illumination, ink bleed-through and blur caused by humidity. Thresholding of such document images either result in broken characters or detection of false texts. Numerous algorithms exist that can separate text and background efficiently in the textual regions of the document; but portions of background are mistaken as text in areas that hardly contain any text. This paper presents a way to overcome these problems by a robust binarization technique that recovers the text from a severely degraded document images and thereby increases the accuracy of optical character recognition systems. The proposed document recovery algorithm efficiently removes degradations from document images. Proposed work is based on the fusion of two well known binarization methods: Gatos et al. and Niblack, using dilation and logical AND operations. The results of our proposed binarization approach are seen to be better when compared to five existing well known approaches proposed by Otsu, Gatos et al., Niblack, Souvola et al., and Bernsen using four evaluations measures: Execution time, F-measure, PSNR, and NRM.

show abstract

“…Furthermore, Trier's evaluation tests [24] identify Niblack's technique, with a post-processing step, as the best. Kamel and Zhao [18] use six evaluation aspects (subjective evaluation, memory, speed, SW restriction, number of parameters of each technique) to evaluate and analyse seven character/graphics extraction based binarisation techniques.…”

Section: Introductionmentioning

confidence: 99%

Document binarisation using Kohonen SOM

Badekas

Papamarkos

2007

IET Image Process.

View full text Add to dashboard Cite

An integrated system for the binarisation of normal and degraded printed documents for the purpose of visualisation and recognition of text characters is proposed. In degraded documents, where considerable background noise or variation in contrast and illumination exists, there are many pixels that cannot be easily classified as foreground or background pixels. For this reason, it is necessary to perform document binarisation by combining and taking into account the results of a set of binarisation techniques, especially for document pixels that have high vagueness. The proposed binarisation technique takes advantages of the benefits of a set of selected binarisation algorithms by combining their results using a Kohonen selforganising map neural network. In order to improve further the binarisation results, significant improvements are proposed for two of the most powerful document binarisation techniques used, that is for the adaptive logical level technique and for the improvement of integrated function algorithm. The proposed binarisation technique is extensively tested with a variety of degraded documents. Several experimental and comparative results, demonstrating the performance of the proposed technique, are presented.

show abstract

Goal-directed evaluation of binarization methods

Cited by 530 publications

References 26 publications

A novel computational approach for automatic dendrite spines detection in two-photon laser scan microscopy

A novel computational approach for automatic dendrite spines detection in two-photon laser scan microscopy

Efficient binarization technique for severely degraded document images

Document binarisation using Kohonen SOM

Contact Info

Product

Resources

About