Colour FAST (CFAST) match: fast affine template matching for colour images

Di, Jia; Cao, Jiang; Song, Weidong; Tang, Xiaoliang; Zhu, Hong

doi:10.1049/el.2016.1331

Cited by 23 publications

(15 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The annual inspection label

I_{1}

is extracted from the query vehicle and regarded as the template while another annual inspection label

I_{2}

is obtained from image in the dataset. As explained in [7], the method to measure similarity distance between

I_{1}

and

I_{2}

is as follows:

Δ_{T} ()(, I_{1}, I_{2})

is the (normalised) sum‐of‐absolute‐difference distance between template

I_{1}

and image

I_{2}

, including a transformation T that maps pixels

p \in I_{1}

to pixels in

I_{2}

[18]. Mathematical formulas can be described as

Δ_{T} false(I_{1}, I_{2} false) = \frac{1}{n_{1}^{2}} \underset{p \in I_{1}}{false\sum} ∥ I_{1} (p) - I_{2} (T (p)) ∥,

furthermore, when the input images

I_{1}

and

I_{2}

are coloured, they need to be converted to greyscale images.…”

Section: Proposed Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Colour fast‐match for precise vehicle retrieval

Wei

Wang

Liu

et al. 2020

J. eng.

View full text Add to dashboard Cite

“…The annual inspection label

I_{1}

is extracted from the query vehicle and regarded as the template while another annual inspection label

I_{2}

is obtained from image in the dataset. As explained in [7], the method to measure similarity distance between

I_{1}

and

I_{2}

is as follows:

Δ_{T} ()(, I_{1}, I_{2})

is the (normalised) sum‐of‐absolute‐difference distance between template

I_{1}

and image

I_{2}

, including a transformation T that maps pixels

p \in I_{1}

to pixels in

I_{2}

[18]. Mathematical formulas can be described as

Δ_{T} false(I_{1}, I_{2} false) = \frac{1}{n_{1}^{2}} \underset{p \in I_{1}}{false\sum} ∥ I_{1} (p) - I_{2} (T (p)) ∥,

furthermore, when the input images

I_{1}

and

I_{2}

are coloured, they need to be converted to greyscale images.…”

Section: Proposed Methodsmentioning

confidence: 99%

“…After greyscale image conversion, the red, green, and blue regions have the same intensity value 23, which results in being undistinguishable from each other in the greyscale image [18]. To solve the problem in Fig.…”

Section: Proposed Methodsmentioning

confidence: 99%

Colour fast‐match for precise vehicle retrieval

Wei

Wang

Liu

et al. 2020

J. eng.

View full text Add to dashboard Cite

“…In general, the TM involves shifting a template obtained from a source image over a search area in an examined image, measuring the similarity between the template and the current search area, and identifying the best candidate image, which matches the template most. As similarity measure methods, the normalized cross correlation (NCC) [22] and sum of absolute difference (SAD) [23] are the two most well-known and widely used in TM applications. Although NCC is more robust than SAD under variable illumination conditions, it needs to compute the numerator and denominator, which is more time-consuming than SAD.…”

Section: Related Workmentioning

confidence: 99%

Generating a Cylindrical Panorama from a Forward-Looking Borehole Video for Borehole Condition Analysis

et al. 2019

View full text Add to dashboard Cite

Geological exploration plays a fundamental and crucial role in geological engineering. The most frequently used method is to obtain borehole videos using an axial view borehole camera system (AVBCS) in a pre-drilled borehole. This approach to surveying the internal structure of a borehole is based on the video playback and video screenshot analysis. One of the drawbacks of AVBCS is that it provides only a qualitative description of borehole information with a forward-looking borehole video, but quantitative analysis of the borehole data, such as the width and dip angle of fracture, are unavailable. In this paper, we proposed a new approach to create a whole borehole-wall cylindrical panorama from the borehole video acquired by AVBCS, which provides a possibility for further analysis of borehole information. Firstly, based on the Otsu and region labeling algorithms, a borehole center location algorithm is proposed to extract the borehole center of each video image automatically. Afterwards, based on coordinate mapping (CM), a virtual coordinate graph (VCG) is designed in the unwrapping process of the front view borehole-wall image sequence, generating the corresponding unfolded image sequence and reducing the computational cost. Subsequently, based on the sum of absolute difference (SAD), a projection transformation SAD (PTSAD), which considers the gray level similarity of candidate images, is proposed to achieve the matching of the unfolded image sequence. Finally, an image filtering module is introduced to filter the invalid frames and the remaining frames are stitched into a complete cylindrical panorama. Experiments on two real-world borehole videos demonstrate that the proposed method can generate panoramic borehole-wall unfolded images from videos with satisfying visual effect for follow up geological condition analysis. From the resulting image, borehole information, including the rock mechanical properties, distribution and width of fracture, fault distribution and seam thickness, can be further obtained and analyzed.

show abstract

“…A significant number of research papers [12][13][14][15][16] have been published in the field of straight line matching. In terms of area feature matching, Fast-Match [17] and CFast-Match [18] are fast algorithms for approximate template matching under 2D affine transformations that minimize the Sum-of-absolutedifferences (SAD) and Colour-sum-of-absolute-differences (CSAD) error measure, and the speed and accuracy of template matching methods have been improved by these methods, In addition, these template matching methods have been applied to many advanced applications [19][20][21][22] of image processing. The shortcoming of the affine matching method is that it dose not work well when nontexture regions are selected as templates.…”

Section: Introductionmentioning

confidence: 99%

“…Bases on template matching methods [17,18] , we deal with the problems from two aspects: 1) Vector sampling normalized cross correlation (VSNCC) is proposed to measure the regional consistency between the two images by multi-channel features, and the effective contrast information among multi-channel features is increased to reduce the impact of noise and illumination in template matching; 2) To improve the accuracy of template matching, three mathematic properties of this method are used in the process of affine transformation to construct texture regions which have affine invariant properties. At last, we prove that our method had a higher accuracy in a case where there was no big distortion between the two viewing angles, the results were verified by simulated images and photographic images.…”

Section: Introductionmentioning

confidence: 99%