New in Image Processing Version 12 continues to utilize modern machine learning and neural networks to add state-of-the-art, high-level, efficient computer vision functions for object detection and recognition, facial analysis, text recognition segmentation and many more. In addition, a large number of new functions and updates have been introduced to enable easier workflow of applications in research and development, computational photography, microscopy, astronomy, medicine and more. Expanded built-in computer vision functions. » Automated object detection and recognition. » Improved face detection and facial feature analysis. » Improved and controllable text recognition. » Built-in image style transfer using one or more templates. » Immediately accessible pre-trained neural nets for images. » Support for image pyramids for multiscale processing. » Specification of color neighborhoods for smooth color segmentation. » Expanded color analysis & processing for image recoloring. » Improved and new color quantization methods. » Automated brightness equalization to correct for uneven illumination. » Conversion of images to vector graphics. » Rasterization of 2D and 3D regions into images. » Addition of new types of image keypoints. » Immediate support for image arithmetic and statistical operations. » Updated the internal storage of image data. » Related Examples Built-in Object Detection & Recognition » Improved Face Detection and Analysis » Large Set of Pre-Trained Neural Nets for Images » Classification of FashionMNIST » ImagePyramid for Multiscale Image Representation » Efficient Zooming Using Multiscale Pyramids » Multiscale Feature Detection » Multiscale Image Blending » Color Neighborhoods » Sophisticated Color Replacement » Improved Color-Based Segmentation » New Color Quantization Methods »
