Dr. Vijayan Asari, DirectorPhone: 937-229-4504
Enhancement of High Contrast Images
The objective of this research is to develop an algorithm which is capable of simultaneous compression of bright regions and enhancement of dark regions in an image frames captured by a video surveillance system. It is also a preprocessing technique to improve the visibility and highlight the features and details of the image for pattern recognition under non-uniform lighting conditions. Two nonlinear image enhancement algorithms - MWIS (Multi-windowed inverse sigmoid), LTSN (locally tuned sine nonlinear) were developed based on image dependent nonlinear functions for enhancing extremely high contrast images. It is observed that MWIS and LTSN algorithms provide extremely optimal results for preprocessing images so that the object regions in an image captured under extremely low or non-uniform lighting conditions are brightened and made more distinct.
The image enhancement algorithms mainly consist of three processes: adaptive intensity enhancement, contrast enhancement, and color restoration. Adaptive intensity enhancement uses a nonlinear transfer function that provides various desired curves that are capable of reducing the intensity of bright pixels while enhancing the intensity of dark pixels in an image at the same time. A control parameter is estimated to adjust the curves for dark and bright pixels. This control parameter is adaptively calculated based on image statistics (Multi-level Gaussian function). Contrast enhancement tunes the intensity of each pixel magnitude based on its surrounding pixels. Finally, a linear color restoration process based on the chromatic information of the input image frame is applied to convert the enhanced intensity image back to a color image.
Some of the enhancement results by processing the color images with the proposed algorithms are provided below and they yield visually optimal results on images captured under extreme lighting conditions. The algorithm would be a promising image enhancement technique that can be useful in many pattern analysis, object detection/recognition and tracking applications.