Color image segmentation using saturated RGB colors and decoupling the intensity from the hue

García-Lamont, Farid; Cervantes, Jair; López-Chau, Asdrúbal; Ruiz-Castilla, Sergio

doi:10.1007/s11042-019-08278-6

Cited by 8 publications

(5 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The training set is built with representative hue samples; for the RGB space, the reference 46 shows that hue samples of the inner faces of the RGB cube are representative enough to train the NN. The training set is built as follows, the elements of the set

normalΘ

are numbers 3

n \in ([], 0, 90)

interval:

normalΘ goodbreak= ({}, 3 n ∣ 0 \leq n \leq 30, n \in normalℤ)

The sets

S

and

C

are built computing the sinus and cosine values of the elements of the set

normalΘ

respectively.

S goodbreak= ({}|, \sin θ_{k}, \forall θ_{k} \in normalΘ)

C goodbreak= ({}|, \cos θ_{k}, \forall θ_{k} \in normalΘ)

Using the Cartesian product, in the sets

P_{1}

P_{2}

, and

P_{3}

we obtain the color vectors of the inner faces of the RGB cube for the planes R‐G, G‐B, and R‐B, respectively.

P_{1} goodbreak= C goodbreak\times S goodbreak\times ({}, 0)

P_{2} goodbreak= ({}, 0) goodbreak\times C goodbreak\times S

P_{3} goodbreak= S goodbreak\times ({}, 0) goodbreak\times C

Finally, the training set

P_{ϕ}

is obtained with:

P_{ϕ} goodbreak= \cup_{i = 1}^{3} P_{i}

While for the HSV space, reference…”

Section: Segmentation Proposalmentioning

confidence: 99%

“…We use the HSV space because the color representation emulates the human perception of color, 43 since the chromaticity and the intensity are decoupled for color processing. [43][44][45][46][47] Note the selection of pixels whose chromatic variance is high regarding the hue of the other elements of the blood smear. The RGB color space is often employed for color processing because of its sensitivity to non-uniform illumination, [44][45][46] despite it is not adequate for color processing; the RGB space can be used because the images are acquired under controlled illumination conditions.…”

Section: Introductionmentioning

confidence: 99%

“…[43][44][45][46][47] Note the selection of pixels whose chromatic variance is high regarding the hue of the other elements of the blood smear. The RGB color space is often employed for color processing because of its sensitivity to non-uniform illumination, [44][45][46] despite it is not adequate for color processing; the RGB space can be used because the images are acquired under controlled illumination conditions. Therefore, we experimented with both color spaces and then compared the results obtained.…”

Section: Introductionmentioning

confidence: 99%

“…We use the HSV space because the color representation emulates the human perception of color, 43 since the chromaticity and the intensity are decoupled for color processing 43‐47 . Note the selection of pixels whose chromatic variance is high regarding the hue of the other elements of the blood smear.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Efficient nucleus segmentation of white blood cells mimicking the human perception of color

García-Lamont

Alvarado²,

López-Chau

et al. 2021

Color Research & Application

Self Cite

View full text Add to dashboard Cite

In this study, we present a nucleus segmentation proposal of white blood cells (WBCs) using chromatic features. It is human inspired on perception of color: a person locates the nucleus of the WBCs by the chromatic contrast between the nucleus and the other elements of the blood smear. To implement that, we segment the nucleus by selecting the pixels with high chromatic variance. First, an unsupervised neural network, which was trained offline to recognize different colors is applied to the images. Thereby, the hue of the pixels is normalized, and the chromatic variance is accurately computed. Processing the hue and using the unsupervised neural network the brightness and staining variations are robustly estimated. In previous related works, the color components are processed separately as uncorrelated intensity channels, and the mathematical operations are selected intuitively. Unlike that, we use color as a feature without separating the hue components, keeping their correlation, so the formal treat becomes systematic. Experiments use the RGB and HSV spaces with three public image databases: ALL‐IDB2, CellaVision, and JTSC. A pixel‐level segmentation evaluation is performed by comparing the segmented images with the ground truth. Our proposal competes with current methods since the values in accuracy, specificity, precision, sensitivity, dice coefficient, kappa index, and true positive rate all are similar to or improved upon the state of the art. The performance of our approach is classified as excellent regarding the kappa index value, and it detects at least 80% of the cells with an average dice coefficient larger than 0.9.

show abstract

normalΘ

are numbers 3

n \in ([], 0, 90)

interval:

normalΘ goodbreak= ({}, 3 n ∣ 0 \leq n \leq 30, n \in normalℤ)

The sets

S

and

C

are built computing the sinus and cosine values of the elements of the set

normalΘ

respectively.

S goodbreak= ({}|, \sin θ_{k}, \forall θ_{k} \in normalΘ)

C goodbreak= ({}|, \cos θ_{k}, \forall θ_{k} \in normalΘ)

Using the Cartesian product, in the sets

P_{1}

P_{2}

, and

P_{3}

we obtain the color vectors of the inner faces of the RGB cube for the planes R‐G, G‐B, and R‐B, respectively.

P_{1} goodbreak= C goodbreak\times S goodbreak\times ({}, 0)

P_{2} goodbreak= ({}, 0) goodbreak\times C goodbreak\times S

P_{3} goodbreak= S goodbreak\times ({}, 0) goodbreak\times C

Finally, the training set

P_{ϕ}

is obtained with:

P_{ϕ} goodbreak= \cup_{i = 1}^{3} P_{i}

While for the HSV space, reference…”

Section: Segmentation Proposalmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Efficient nucleus segmentation of white blood cells mimicking the human perception of color

García-Lamont

Alvarado²,

López-Chau

et al. 2021

Color Research & Application

Self Cite

View full text Add to dashboard Cite

show abstract

“…So, it is required to optimize the image provided for use in the carpet map design. This optimization includes a reduction in size and number of hues in the original color images or wool fibers 9‐11 . Currently, experts perform this operation manually using general software such as Photoshop.…”

Section: Introductionmentioning

confidence: 99%

A hybrid method for reduction of size and number of hues in the color images used in a carpet map

Cui

Shi

2021

Color Research & Application

View full text Add to dashboard Cite

The carpet industry is no longer a small business done in the villages on a small scale; instead, it has carved an altogether different space, identity, and appreciation for itself in the cosmopolitan world. As computers are is becoming more and more ubiquitous, most industries, including the carpet industry, use computers for quality improvement, accuracy enhancement, speed development, and cost reduction purposes. Unlike traditional carpet maps, many modern maps include images of human faces for hand‐woven carpet tableau. These digital images comprise millions of colors and thousands of pixels, making it practically impossible to construct and weave the carpet in the same dimensions. Many weavers currently use manual and experience‐based methods for reducing the size and number of hues for making a hand‐woven carpet tableau map. Therefore, the outcomes are not the optimal results and can be improved. Also, many color reduction methods do not focus on the hand‐woven carpet tableau map. To overcome these problems and gaps, this research focuses on proposing a new automatic method for reducing the size of color images without compromising facial nuances, lessening the number of colors used while protecting the important areas of the images, and transforming those images into carpet tableau maps. The proposed approach inputs the original color image. It continuously detecting the face and specifying important areas, and finally, outputs carpet tableau map that is proportional to the given dimensions and color count. To evaluate the proposed method, MATLAB, as a powerful simulation tool, was employed. Final results are compared to the existing approaches in terms of face detection, size reduction, and color quantization. The obtained results have shown that the approach improves speed by 39% in face detection and increases the precision of size reduction and color quantization phases. The results have also confirmed that when images of human faces are reduced by a proposed method to form an appropriate image for tableau maps, they are nearly always perceived as more attractive than the reduced faces via traditional methods.

show abstract

A Forensic Video Upscaling Colorizing and Denoising Framework for Crime Scene Investigation

Prema,

Anita

2023

Lecture Notes in Electrical Engineering

View full text Add to dashboard Cite

Color image segmentation using saturated RGB colors and decoupling the intensity from the hue

Cited by 8 publications

References 51 publications

Efficient nucleus segmentation of white blood cells mimicking the human perception of color

Efficient nucleus segmentation of white blood cells mimicking the human perception of color

A hybrid method for reduction of size and number of hues in the color images used in a carpet map

A Forensic Video Upscaling Colorizing and Denoising Framework for Crime Scene Investigation

Contact Info

Product

Resources

About