{"id":264,"date":"2016-06-17T13:51:10","date_gmt":"2016-06-17T11:51:10","guid":{"rendered":"http:\/\/zaven.co\/blog\/?p=264"},"modified":"2025-04-08T19:55:20","modified_gmt":"2025-04-08T17:55:20","slug":"opencv-advanced-android-development-light-source-detection","status":"publish","type":"post","link":"https:\/\/zaven.co\/blog\/opencv-advanced-android-development-light-source-detection\/","title":{"rendered":"OpenCV in advanced Android development: light source detection"},"content":{"rendered":"

In the last Android development tutorial we\u2019ve covered drawing a histogram using OpenCV<\/a>. This time we\u2019ll show you how to detect light sources in an image<\/strong>. This feature comes in very handy when building an Android app that has to perform quality checks on user-generated photographs (e.g. sending documents for proof).<\/p>\n

Our goal is to find the light source point and draw there a circle in an Android app<\/strong>. The easiest way to get it done is to detect the localization where the greatest white pixel value is. We can do it by using the Core.minMaxLoc<\/code> method<\/strong>, which takes only one parameter – the array.<\/p>\n

In order to detect light source with OpenCV in our Android application<\/em> we need to perform 4 steps:<\/p>\n

    \n
  1. Read the bitmap<\/li>\n
  2. Transform it to grayscale<\/li>\n
  3. Apply Gaussian blur<\/li>\n
  4. Calculate maxLoc<\/code> by using the minMaxLoc<\/code> method<\/li>\n<\/ol>\n
    \n    private void detectLight(Bitmap bitmap, double gaussianBlurValue) {\n        Mat rgba = new Mat();\n        Utils.bitmapToMat(bitmap, rgba);\n\n        Mat grayScaleGaussianBlur = new Mat();\n        Imgproc.cvtColor(rgba, grayScaleGaussianBlur, Imgproc.COLOR_BGR2GRAY);\n        Imgproc.GaussianBlur(grayScaleGaussianBlur, grayScaleGaussianBlur, new Size(gaussianBlurValue, gaussianBlurValue), 0);\n\n        Core.MinMaxLocResult minMaxLocResultBlur = Core.minMaxLoc(grayScaleGaussianBlur);\n        Imgproc.circle(rgba, minMaxLocResultBlur.maxLoc, 30, new Scalar(255), 3);\n\n        \/\/ Don't do that at home or work it's for visualization purpose.\n        Bitmap resultBitmap = Bitmap.createBitmap(rgba.cols(), rgba.rows(), Bitmap.Config.ARGB_8888);\n        Utils.matToBitmap(rgba, resultBitmap);\n        BitmapHelper.showBitmap(this, resultBitmap, detectLightImageView);\n\n        Bitmap blurryBitmap = Bitmap.createBitmap(grayScaleGaussianBlur.cols(), grayScaleGaussianBlur.rows(), Bitmap.Config.ARGB_8888);\n        Utils.matToBitmap(grayScaleGaussianBlur, blurryBitmap);\n        BitmapHelper.showBitmap(this, blurryBitmap, gaussianBlurImageView);\n\n    }\n<\/code><\/pre>\n
    Extract data and apply transformations<\/h2>\n
    \/\/ Covert our bitmap into an array\nMat rgba = new Mat();\nUtils.bitmapToMat(bitmap, rgba);\n\n\/\/ Convert our rgba array to greyscale by using Imgproc.cvtColor ()\n\/\/ Apply Gaussian blur (for the last one we recommend Imgproc.GaussianBlur()).\nMat grayScaleGaussianBlur = new Mat();\nImgproc.cvtColor(rgba, grayScaleGaussianBlur, Imgproc.COLOR_BGR2GRAY);\nImgproc.GaussianBlur(grayScaleGaussianBlur, grayScaleGaussianBlur, new Size(gaussianBlurValue, gaussianBlurValue), 0);\n<\/code><\/pre>\n
    Imgproc.cvtColor<\/code> parameters:<\/p>\n