From: Greg Burri Date: Fri, 18 Dec 2015 19:27:24 +0000 (+0100) Subject: Add functions to apply an area opening and to compute granulometry. X-Git-Tag: 1.0.11~78 X-Git-Url: http://git.euphorik.ch/index.cgi?a=commitdiff_plain;h=b070295cf67b2025164a34b6594e84f0d771cdc9;p=master-thesis.git Add functions to apply an area opening and to compute granulometry. --- diff --git a/Parasitemia/Parasitemia/Classifier.fs b/Parasitemia/Parasitemia/Classifier.fs index 8467d24..3ed42f0 100644 --- a/Parasitemia/Parasitemia/Classifier.fs +++ b/Parasitemia/Parasitemia/Classifier.fs @@ -40,13 +40,13 @@ let findCells (ellipses: Ellipse list) (parasites: ParasitesMarker.Result) (fg: let a = int (e.A + 0.5) cx - a, cy - a, cx + a, cy + a - // 1) Remove ellipses touching the edges. let w = float fg.Width let h = float fg.Height - let ellipsesInside = ellipses |> List.map (fun e -> EllipseFlaggedKd (e, Removed = e.isOutside w h)) - // 2) Associate touching ellipses with each ellipses. - let tree = KdTree.Tree.BuildTree ellipsesInside + let ellipses = ellipses |> List.map EllipseFlaggedKd + + // 1) Associate touching ellipses with each ellipses. + let tree = KdTree.Tree.BuildTree ellipses let neighbors (e: Ellipse) : (EllipseFlaggedKd * PointD * PointD) list = tree.Search (searchRegion e) // We only keep the ellipses touching 'e'. @@ -57,10 +57,10 @@ let findCells (ellipses: Ellipse list) (parasites: ParasitesMarker.Result) (fg: | _ -> None ) - let ellipsesWithNeigbors = ellipsesInside |> List.choose (fun e -> if e.Removed then None else Some (e, neighbors e)) + let ellipsesWithNeigbors = ellipses |> List.choose (fun e -> if e.Removed then None else Some (e, neighbors e)) - // 3) Remove ellipses with a lower percentage of foreground. - for e, neighbors in ellipsesWithNeigbors do + // 2) Remove ellipses with a lower percentage of foreground. (taken from the lower score to the highest). + for e, neighbors in List.rev ellipsesWithNeigbors do let minX, minY, maxX, maxY = ellipseWindow e let mutable totalElement = 0 @@ -68,7 +68,7 @@ let findCells (ellipses: Ellipse list) (parasites: ParasitesMarker.Result) (fg: for y in (if minY < 0 then 0 else minY) .. (if maxY >= fg.Height then fg.Height - 1 else maxY) do for x in (if minX < 0 then 0 else minX) .. (if maxX >= fg.Width then fg.Width - 1 else maxX) do let yf, xf = float y, float x - if e.Contains xf yf && neighbors |> List.forall (fun (otherE, _, _) -> not <| otherE.Contains xf yf) + if e.Contains xf yf && neighbors |> List.forall (fun (otherE, _, _) -> otherE.Removed || not <| otherE.Contains xf yf) then totalElement <- totalElement + 1 if fg.Data.[y, x, 0] > 0uy @@ -79,6 +79,10 @@ let findCells (ellipses: Ellipse list) (parasites: ParasitesMarker.Result) (fg: then e.Removed <- true + // 3) Remove ellipses touching the edges. + for e in ellipses do + if e.isOutside w h then e.Removed <- true + // 4) Remove ellipses with little area. for e, neighbors in ellipsesWithNeigbors do if not e.Removed diff --git a/Parasitemia/Parasitemia/Ellipse.fs b/Parasitemia/Parasitemia/Ellipse.fs index 128c823..139b7b5 100644 --- a/Parasitemia/Parasitemia/Ellipse.fs +++ b/Parasitemia/Parasitemia/Ellipse.fs @@ -46,8 +46,8 @@ let private goldenSectionSearch (f: float -> float) (nbIter: int) (xmin: float) // Ellipse.A is always equal or greater than Ellipse.B. // Ellipse.Alpha is between 0 and Pi. let ellipse (p1x: float) (p1y: float) (m1: float) (p2x: float) (p2y: float) (m2: float) (p3x: float) (p3y: float) : Types.Ellipse option = - let accuracy_extremum_search_1 = 4 - let accuracy_extremum_search_2 = 3 + let accuracy_extremum_search_1 = 7 // 3 + let accuracy_extremum_search_2 = 7 // 4 // p3 as the referencial. let p1x = p1x - p3x @@ -178,18 +178,31 @@ let private areVectorsValid (p1x: float) (p1y: float) (p2x: float) (p2y: float) let b1 = -m1 * p1x + p1y let b2 = -m2 * p2x + p2y - let px = -((b1 - b2)/(m1 - m2)) - let py = -((m2 * b1 - m1 * b2)/(m1 - m2)) + let px = -((b1 - b2) / (m1 - m2)) + let py = -((m2 * b1 - m1 * b2) / (m1 - m2)) let rot1 = vectorRotation p1x p1y v1x v1y px py let rot2 = vectorRotation p2x p2y v2x v2y px py - if rot1 = rot2 || rot1 * atan2 (p1y - py) (p1x - px) + rot2 * atan2 (p2y - py) (p2x - px) <= 0.0 + if rot1 = rot2 then None else + let alpha1 = atan2 (p1y - py) (p1x - px) + let alpha2 = atan2 (p2y - py) (p2x - px) + + let alpha1' = if alpha1 < 0.0 then 2.0 * Math.PI + alpha1 else alpha1 + let alpha2' = if alpha2 < 0.0 then 2.0 * Math.PI + alpha2 else alpha2 + + let diff = rot1 * alpha1' + rot2 * alpha2' + + if diff > Math.PI || (diff < 0.0 && diff > -Math.PI) + then + None + else Some (m1, m2) + let find (edges: Matrix) (xDir: Image) (yDir: Image) @@ -262,6 +275,8 @@ let find (edges: Matrix) match ellipse p1xf p1yf m1 p2xf p2yf m2 p3xf p3yf with | Some e when e.Cx > 0.0 && e.Cx < (float w) - 1.0 && e.Cy > 0.0 && e.Cy < (float h) - 1.0 && e.A >= r1 - radiusTolerance && e.A <= r2 + radiusTolerance && e.B >= r1 - radiusTolerance && e.B <= r2 + radiusTolerance -> + + let prout = areVectorsValid p1xf p1yf p2xf p2yf -xDirData.[p1y, p1x, 0] -yDirData.[p1y, p1x, 0] -xDirData.[p2y, p2x, 0] -yDirData.[p2y, p2x, 0] ellipses.Add e | _ -> () | _ -> () diff --git a/Parasitemia/Parasitemia/Granulometry.fs b/Parasitemia/Parasitemia/Granulometry.fs new file mode 100644 index 0000000..a9a102a --- /dev/null +++ b/Parasitemia/Parasitemia/Granulometry.fs @@ -0,0 +1,38 @@ +module Granulometry + +open System +open System.Drawing + +open Emgu.CV +open Emgu.CV.Structure + +open Utils + +// 'range': a minimum and maximum radius. +// 'scale': <= 1.0, to speed up the process. +let findRadius (img: Image) (range: int * int) (scale: float) : int = + use scaledImg = if scale = 1.0 then img else img.Resize(scale, CvEnum.Inter.Area) + + let r1, r2 = range + let r1', r2' = roundInt (float r1 * scale), roundInt (float r2 * scale) + + let patternSpectrum = Array.zeroCreate (r2' - r1') + let intensityImg = scaledImg.GetSum().Intensity + + let mutable previous_n = Double.NaN + for r in r1' .. r2' do + let se = CvInvoke.GetStructuringElement(CvEnum.ElementShape.Ellipse, Size(2 * r, 2 * r), Point(-1, -1)) + use closed = scaledImg.MorphologyEx(CvEnum.MorphOp.Close, se, Point(-1, -1), 1, CvEnum.BorderType.Replicate, MCvScalar(0.0)) + + let n = 1.0 - closed.GetSum().Intensity / intensityImg + + if not (Double.IsNaN previous_n) + then + patternSpectrum.[r - r1' - 1] <- abs (n - previous_n) + previous_n <- n + + let max = patternSpectrum |> Array.indexed |> Array.fold (fun (iMax, sMax) (i, s) -> if s > sMax then (i, s) else (iMax, sMax)) (0, Double.MinValue) + + 0 + + diff --git a/Parasitemia/Parasitemia/ImgTools.fs b/Parasitemia/Parasitemia/ImgTools.fs index 7253390..e9cbbe9 100644 --- a/Parasitemia/Parasitemia/ImgTools.fs +++ b/Parasitemia/Parasitemia/ImgTools.fs @@ -3,11 +3,13 @@ open System open System.Drawing open System.Collections.Generic +open System.Linq open Emgu.CV open Emgu.CV.Structure open Utils +open Heap // Normalize image values between 0uy and 255uy. let normalizeAndConvert (img: Image) : Image = @@ -18,10 +20,266 @@ let normalizeAndConvert (img: Image) : Image = img.MinMax(min, max, minLocation, maxLocation) ((img - (!min).[0]) / ((!max).[0] - (!min).[0]) * 255.0).Convert() + let gaussianFilter (img : Image<'TColor, 'TDepth>) (standardDeviation : float) : Image<'TColor, 'TDepth> = let size = 2 * int (ceil (4.0 * standardDeviation)) + 1 img.SmoothGaussian(size, size, standardDeviation, standardDeviation) + +let findMaxima (img: Image) : IEnumerable> = + use suppress = new Image(img.Size) + let w = img.Width + let h = img.Height + + let imgData = img.Data + let suppressData = suppress.Data + + let result = List>() + + let flood (start: Point) : List> = + let sameLevelToCheck = Stack() + let betterLevelToCheck = Stack() + betterLevelToCheck.Push(start) + + let result' = List>() + + while betterLevelToCheck.Count > 0 do + let p = betterLevelToCheck.Pop() + if suppressData.[p.Y, p.X, 0] = 0uy + then + suppressData.[p.Y, p.X, 0] <- 1uy + sameLevelToCheck.Push(p) + let current = List() + + let mutable betterExists = false + + while sameLevelToCheck.Count > 0 do + let p' = sameLevelToCheck.Pop() + let currentLevel = imgData.[p'.Y, p'.X, 0] + current.Add(p') |> ignore + for i in -1 .. 1 do + for j in -1 .. 1 do + if i <> 0 || j <> 0 + then + let ni = i + p'.Y + let nj = j + p'.X + if ni >= 0 && ni < h && nj >= 0 && nj < w + then + let level = imgData.[ni, nj, 0] + let notSuppressed = suppressData.[ni, nj, 0] = 0uy + + if level = currentLevel && notSuppressed + then + suppressData.[ni, nj, 0] <- 1uy + sameLevelToCheck.Push(Point(nj, ni)) + elif level > currentLevel + then + betterExists <- true + if notSuppressed + then + betterLevelToCheck.Push(Point(nj, ni)) + + if not betterExists + then + result'.Add(current) + result' + + for i in 0 .. h - 1 do + for j in 0 .. w - 1 do + let maxima = flood (Point(j, i)) + if maxima.Count > 0 + then result.AddRange(maxima) + + result.Select(fun l -> HashSet(l)) + + +type PriorityQueue () = + let q = List>() // TODO: Check performance with an HasSet + let mutable highest = -1 // Value of the first elements of 'q'. + + member this.Next () : byte * Point = + if this.IsEmpty + then + invalidOp "Queue is empty" + else + let l = q.[0] + let next = l.First() + l.Remove(next) |> ignore + let value = byte highest + if l.Count = 0 + then + q.RemoveAt(0) + highest <- highest - 1 + while q.Count > 0 && q.[0] = null do + q.RemoveAt(0) + highest <- highest - 1 + value, next + + member this.Max = + highest |> byte + + (*member this.UnionWith (other: PriorityQueue) = + while not other.IsEmpty do + let p, v = other.Next + this.Add p v*) + + member this.Add (value: byte) (p: Point) = + let vi = int value + + if this.IsEmpty + then + highest <- int value + q.Insert(0, null) + elif vi > highest + then + for i in highest .. vi - 1 do + q.Insert(0, null) + highest <- vi + elif highest - vi >= q.Count + then + for i in 0 .. highest - vi - q.Count do + q.Add(null) + + let pos = highest - vi + if q.[pos] = null + then + q.[pos] <- HashSet([p]) + else + q.[pos].Add(p) |> ignore + + + member this.IsEmpty = + q.Count = 0 + + member this.Clear () = + while highest >= 0 do + q.[highest] <- null + highest <- highest - 1 + + + +type MaximaState = Uncertain | Validated | TooBig +type Maxima = { + elements : HashSet + mutable intensity: byte option + mutable state: MaximaState } + + +let areaOpen (img: Image) (area: int) = + let w = img.Width + let h = img.Height + + let maxima = findMaxima img |> Seq.map (fun m -> { elements = m; intensity = None; state = Uncertain }) |> List.ofSeq + let toValidated = Stack(maxima) + + while toValidated.Count > 0 do + let m = toValidated.Pop() + if m.elements.Count <= area + then + let queue = + let q = PriorityQueue() + let firstElements = HashSet() + for p in m.elements do + for i in -1 .. 1 do + for j in -1 .. 1 do + if i <> 0 || j <> 0 + then + let ni = i + p.Y + let nj = j + p.X + let p' = Point(nj, ni) + if ni >= 0 && ni < h && nj >= 0 && nj < w && not (m.elements.Contains(p')) && not (firstElements.Contains(p')) + then + firstElements.Add(p') |> ignore + q.Add (img.Data.[ni, nj, 0]) p' + q + + let mutable intensity = queue.Max + let nextElements = HashSet() + + let mutable stop = false + while not stop do + let intensity', p = queue.Next () + + if intensity' = intensity // The intensity doesn't change. + then + if m.elements.Count + nextElements.Count + 1 > area + then + m.state <- Validated + m.intensity <- Some intensity + stop <- true + else + nextElements.Add(p) |> ignore + elif intensity' < intensity + then + m.elements.UnionWith(nextElements) + if m.elements.Count = area + then + m.state <- Validated + m.intensity <- Some (intensity') + stop <- true + else + intensity <- intensity' + nextElements.Clear() + nextElements.Add(p) |> ignore + else // i' > i + seq { + for m' in maxima do + if m' <> m && m'.elements.Contains(p) then + if m'.elements.Count + m.elements.Count <= area + then + m'.state <- Uncertain + m'.elements.UnionWith(m.elements) + if not <| toValidated.Contains m' // FIXME: Maybe use state instead of scanning the whole list. + then + toValidated.Push(m') + stop <- true + yield false + } |> Seq.forall id |> ignore + + if not stop + then + m.state <- Validated + m.intensity <- Some (intensity) + stop <- true + + if not stop + then + for i in -1 .. 1 do + for j in -1 .. 1 do + if i <> 0 || j <> 0 + then + let ni = i + p.Y + let nj = j + p.X + let p' = Point(nj, ni) + if ni < 0 || ni >= h || nj < 0 || nj >= w + then + m.state <- Validated + m.intensity <- Some (intensity) + stop <- true + elif not (m.elements.Contains(p')) && not (nextElements.Contains(p')) + then + queue.Add (img.Data.[ni, nj, 0]) p' + + if queue.IsEmpty + then + if m.elements.Count + nextElements.Count <= area + then + m.state <- Validated + m.intensity <- Some intensity' + m.elements.UnionWith(nextElements) + stop <- true + + for m in maxima do + if m.state = Validated + then + match m.intensity with + | Some i -> + for p in m.elements do + img.Data.[p.Y, p.X, 0] <- i + | _ -> () + () + + // Zhang and Suen algorithm. // Modify 'mat' in place. let thin (mat: Matrix) = @@ -73,6 +331,7 @@ let thin (mat: Matrix) = data2 <- tmp +// FIXME: replace by a queue or stack. let pop (l: List<'a>) : 'a = let n = l.[l.Count - 1] l.RemoveAt(l.Count - 1) diff --git a/Parasitemia/Parasitemia/KMedians.fs b/Parasitemia/Parasitemia/KMedians.fs index 1822c75..9b5d50f 100644 --- a/Parasitemia/Parasitemia/KMedians.fs +++ b/Parasitemia/Parasitemia/KMedians.fs @@ -10,7 +10,7 @@ type Result = { fg: Image median_bg: float median_fg: float - d_fg: Image } // Distances to median_fg. + d_fg: Image } // Euclidean distances of the foreground to median_fg. let kmedians (img: Image) (fgFactor: float) : Result = let nbIteration = 3 @@ -32,17 +32,19 @@ let kmedians (img: Image) (fgFactor: float) : Result = for i in 1 .. nbIteration do CvInvoke.Pow(img - median_bg, 2.0, d_bg) CvInvoke.Pow(img - median_fg, 2.0, d_fg) - fg <- (d_fg * fgFactor).Cmp(d_bg, CvEnum.CmpType.LessThan) + CvInvoke.Compare(d_fg, d_bg, fg, CvEnum.CmpType.LessThan) - median_fg <- MathNet.Numerics.Statistics.Statistics.Median(seq { - for i in 0 .. h - 1 do - for j in 0 .. w - 1 do - if fg.Data.[i, j, 0] > 0uy then yield img.Data.[i, j, 0] |> float }) + let bg_values = List() + let fg_values = List() - median_bg <- MathNet.Numerics.Statistics.Statistics.Median(seq { - for i in 0 .. h - 1 do - for j in 0 .. w - 1 do - if fg.Data.[i, j, 0] = 0uy then yield img.Data.[i, j, 0] |> float }) + for i in 0 .. h - 1 do + for j in 0 .. w - 1 do + if fg.Data.[i, j, 0] > 0uy + then fg_values.Add(float img.Data.[i, j, 0]) + else bg_values.Add(float img.Data.[i, j, 0]) + + median_bg <- MathNet.Numerics.Statistics.Statistics.Median(bg_values) + median_fg <- MathNet.Numerics.Statistics.Statistics.Median(fg_values) CvInvoke.Sqrt(d_fg, d_fg) diff --git a/Parasitemia/Parasitemia/MainAnalysis.fs b/Parasitemia/Parasitemia/MainAnalysis.fs index 3f35fb7..155bfcc 100644 --- a/Parasitemia/Parasitemia/MainAnalysis.fs +++ b/Parasitemia/Parasitemia/MainAnalysis.fs @@ -14,12 +14,44 @@ open Types let doAnalysis (img: Image) (name: string) (config: Config) : Cell list = - let imgFloat = img.Convert() - use scaledImg = if config.scale = 1.0 then imgFloat else imgFloat.Resize(config.scale, CvEnum.Inter.Area) + use scaledImg = if config.scale = 1.0 then img else img.Resize(config.scale, CvEnum.Inter.Area) + use blue = scaledImg.Item(0) use green = scaledImg.Item(1) + use red = scaledImg.Item(2) - use filteredGreen = (gaussianFilter green config.doGSigma1) - config.doGLowFreqPercentageReduction * (gaussianFilter green config.doGSigma2) + + let greenFloat = green.Convert() + + let green = gaussianFilter green 1.5 + + // let RBCSize = Granulometry.findRadius green (10, 100) 0.5 + + match config.debug with + | DebugOn output -> + let dirPath = System.IO.Path.Combine(output, name) + System.IO.Directory.CreateDirectory dirPath |> ignore + let buildFileName postfix = System.IO.Path.Combine(dirPath, name + postfix) + + saveImg green (buildFileName " - green.png") + + let greenMaxima = green.Copy() + let maxima = ImgTools.findMaxima greenMaxima + for m in maxima do + for p in m do + greenMaxima.Data.[p.Y, p.X, 0] <- 255uy + + saveImg greenMaxima (buildFileName " - maxima.png") + + logTime "areaOpen" (fun () -> ImgTools.areaOpen green 800) + saveImg green (buildFileName " - green opened.png") + + | _ -> () + + [] + (* + + use filteredGreen = (gaussianFilter greenFloat config.doGSigma1) - config.doGLowFreqPercentageReduction * (gaussianFilter greenFloat config.doGSigma2) use sobelKernel = new ConvolutionKernelF(array2D [[ 1.0f; 0.0f; -1.0f ] @@ -61,7 +93,7 @@ let doAnalysis (img: Image) (name: string) (config: Config) : Cell li let kmediansResults = logTime "Finding foreground (k-medians)" (fun () -> KMedians.kmedians filteredGreen 1.0) - let parasites = ParasitesMarker.find green filteredGreen kmediansResults config + let parasites = ParasitesMarker.find greenFloat filteredGreen kmediansResults config let allEllipses, ellipses = logTime "Finding ellipses" (fun () -> let matchingEllipses = Ellipse.find edges xEdges yEdges config @@ -72,7 +104,11 @@ let doAnalysis (img: Image) (name: string) (config: Config) : Cell li // Output pictures if debug flag is set. match config.debug with | DebugOn output -> - let buildFileName postfix = System.IO.Path.Combine(output, name + postfix) + let dirPath = System.IO.Path.Combine(output, name) + System.IO.Directory.CreateDirectory dirPath |> ignore + + let buildFileName postfix = System.IO.Path.Combine(dirPath, name + postfix) + saveMat (edges * 255.0) (buildFileName " - edges.png") saveImg parasites.darkStain (buildFileName " - parasites - dark stain.png") @@ -96,6 +132,11 @@ let doAnalysis (img: Image) (name: string) (config: Config) : Cell li let imgCells' = img.Copy() drawCells imgCells' true cells saveImg imgCells' (buildFileName " - cells - full.png") + + saveImg (normalizeAndConvert filteredGreen) (buildFileName " - filtered.png") + saveImg blue (buildFileName " - blue.png") + saveImg green (buildFileName " - green.png") + saveImg red (buildFileName " - red.png") | _ -> () - cells + cells*) diff --git a/Parasitemia/Parasitemia/MatchingEllipses.fs b/Parasitemia/Parasitemia/MatchingEllipses.fs index da74c65..92a767a 100644 --- a/Parasitemia/Parasitemia/MatchingEllipses.fs +++ b/Parasitemia/Parasitemia/MatchingEllipses.fs @@ -9,17 +9,23 @@ open Types open Utils +// Do not take in account matching score below this when two ellipses are matched. let matchingScoreThreshold1 = 0.6 -let matchingScoreThreshold2 = 1. + +// All ellipsee with a score below this is removed. +let matchingScoreThreshold2 = 1. / 50. type private EllipseScoreFlaggedKd (matchingScore: float, e: Ellipse) = let mutable matchingScore = matchingScore + let perimeter = e.Perimeter member this.Ellipse = e member this.MatchingScore = matchingScore + + // The score is proportional to the perimeter because large ellipse will receive more votes. member this.AddMatchingScore(score: float) = - matchingScore <- matchingScore + score + matchingScore <- matchingScore + score / perimeter member val Processed = false with get, set member val Removed = false with get, set @@ -38,6 +44,7 @@ type MatchingEllipses (radiusMin: float) = member this.Ellipses : Ellipse list = List.ofSeq ellipses |> List.map (fun e -> e.Ellipse) + // Process all ellipses and return ellipses ordered from the best score to the worst. member this.PrunedEllipses : Ellipse list = if ellipses.Count = 0 then @@ -93,7 +100,7 @@ type MatchingEllipses (radiusMin: float) = for other in tree.Search window do if not other.Removed && other.MatchingScore < e.MatchingScore then - if e.Ellipse.Contains other.Ellipse.Cx other.Ellipse.Cy + if e.Ellipse.Scale(0.8).Contains other.Ellipse.Cx other.Ellipse.Cy then other.Removed <- true ellipses.RemoveAll(fun e -> e.Removed) |> ignore diff --git a/Parasitemia/Parasitemia/Parasitemia.fsproj b/Parasitemia/Parasitemia/Parasitemia.fsproj index b9316f1..3e6609a 100644 --- a/Parasitemia/Parasitemia/Parasitemia.fsproj +++ b/Parasitemia/Parasitemia/Parasitemia.fsproj @@ -28,7 +28,7 @@ x64 bin\Debug\Parasitemia.XML false - --folder "../../../../Images/08.10.2015/debug" --output "../../../Images/output" --debug + --folder "../../../Images/areaopen" --output "../../../Images/output" --debug pdbonly @@ -40,7 +40,7 @@ x64 bin\Release\Parasitemia.XML false - --folder "../../../../Images/08.10.2015/release" --output "../../../Images/output" --debug + --folder "../../../Images/release" --output "../../../Images/output" --debug 11 @@ -64,9 +64,11 @@ + + diff --git a/Parasitemia/Parasitemia/Program.fs b/Parasitemia/Parasitemia/Program.fs index d7348cf..fdbc908 100644 --- a/Parasitemia/Parasitemia/Program.fs +++ b/Parasitemia/Parasitemia/Program.fs @@ -65,14 +65,14 @@ let main args = // RBC size range in px at scale = 1.0. minRBCSize = 20. - maxRBCSize = 40. + maxRBCSize = 42. doGSigma1 = 1.5 - doGSigma2 = 20. + doGSigma2 = 30. doGLowFreqPercentageReduction = 0.75 - factorNbPick = 2.0 - factorWindowSize = 1.6 + factorNbPick = 1.0 + factorWindowSize = 2.0 darkStainLevel = 0.4 // Lower -> more sensitive. @@ -88,18 +88,18 @@ let main args = MaxDarkStainRatio = 0.1 - minimumCellArea = 1200. * scale ** 2. |> int + minimumCellArea = 1000. * scale ** 2. |> int maxOffcenter = 0.5 } match mode with | CmdLine (input, output) -> - let config = { config with debug = DebugOn output } + let config = if debug then { config with debug = DebugOn output } else config Directory.CreateDirectory output |> ignore use logFile = new StreamWriter(new FileStream(Path.Combine(output, "log.txt"), FileMode.Append, FileAccess.Write)) Utils.log <- (fun m -> logFile.WriteLine(m)) - Utils.log (sprintf "=== New run : %A ===" DateTime.Now) + Utils.log (sprintf "=== New run : %A %A ===" DateTime.Now (if debug then "[DEBUG]" else "[RELEASE]")) let files = match input with | File file -> [ file ] @@ -111,7 +111,8 @@ let main args = try use img = new Image(file) Utils.log (sprintf "== File: %A" file) - let cells = Utils.logTime "Whole analyze" (fun () -> ImageAnalysis.doAnalysis img (FileInfo(file).Name) config) + + let cells = Utils.logTime "Whole analyze" (fun () -> ImageAnalysis.doAnalysis img (Path.GetFileNameWithoutExtension(FileInfo(file).Name)) config) let total, infected = Utils.countCells cells fprintf resultFile "File: %s %d %d %.2f\n" file total infected (100. * (float infected) / (float total)) with diff --git a/Parasitemia/Parasitemia/Types.fs b/Parasitemia/Parasitemia/Types.fs index d7bacbd..a551142 100644 --- a/Parasitemia/Parasitemia/Types.fs +++ b/Parasitemia/Parasitemia/Types.fs @@ -30,6 +30,13 @@ type Ellipse (cx: float, cy: float, a: float, b: float, alpha: float) = this.CutAVericalLine 0.0 || this.CutAVericalLine width || this.CutAnHorizontalLine 0.0 || this.CutAnHorizontalLine height + member this.Scale (factor: float) = + Ellipse(this.Cx, this.Cy, this.A * factor, this.B * factor, alpha) + + // Approximation of Ramanujan. + member this.Perimeter = + Math.PI * (3.0 * (this.A + this.B) - sqrt ((3.0 * this.A + this.B) * (this.A + 3.0 * this.B))) + type CellClass = HealthyRBC | InfectedRBC | Peculiar