Research On Video Anti-Blooming

Light-blooming occurs when the brightness ratio between the bright and gray parts of video images is beyond the dynamic scope of image sensors. Based on phenomena of light-blooming produced by vehicles traveling during night, the mechanism of light-blooming phenomena is being researched and analyzed in this thesis. Aiming at solving the problem of light-blooming of video images, the author proposes three new approaches of anti-blooming technology and makes further research on realization of them in details.The first anti-blooming approach is that the integration time for each pixel sensitive unit in the CCD image sensor is controlled independently. Based on the current array CCD image sensor technology, the integration time for each pixel sensitive unit in the CCD image sensor can be controlled independently by way of adding circuit for pixel sensitive unit addressing and integration pulse generating. Anti-blooming can be achieved if the integration time for a certain pixel sensitive unit is shortened appropriately when charges overflow potential well of a pixel sensitive unit in the array CCD Image Sensor.The Second anti-blooming approach is that the integration time of each pixel sensitive unit in the CMOS image sensor is controlled independently. On the basis of current array CMOS image sensor structure, the integration time for each pixel sensitive unit in the CMOS image sensor can be controlled independently by way of adding circuits for reset pulse signal generation and reset pulse latches. Anti-blooming can be achieved if the reset pulse width of the pixel sensitive unit is extended appropriately when blooming happened at one pixel sensitive unit in the array CMOS Image Sensor.The third anti-blooming approach is based on two channel CCD image sensor. Two channel image signals of the same scenery can be obtained by two CCD synchronous cameras whose integration times are different:one is longer than another. The CCD camera with short integration time is used for obtaining bright image information while the one with long integration time is used for obtaining gray image information. Two methods might be adopted for images processing:one is to process the images obtained from the two channels through the calculating method for pixel brightness exchange by applying CPLD, and another is to adopt DSP in processing the images obtained from the two channels by way of combining the sub-section linear weight. Then all the processed image information will be input into a video encoder and finally come to one channel video signal with width dynamic range so as to achieve the purpose of anti-blooming.Both the first and second anti-blooming approaches achieve the purpose of anti-blooming by changing the internal structures of image sensors. As to the third approach, anti-blooming is achieved, by using DSP as hardware platform for image processing and applying relevant calculating methods for image processing.It is proved through experiment that the first and second approaches can achieve a certain effect in anti-blooming while the dynamic range of CCD image sensor could be expended from 80dB to 106dB by applying the anti-blooming technology based on the two channel CCD image sensor, which can effectively settle the problem of light-blooming in video images.