US 20060268117 A1
A method for simultaneously recording motion and still images, includes the steps of: capturing a motion image sequence and accompanying audio of a scene with a digital video camera adapted to record both motion and higher resolution still images; simultaneously capturing a still image sequence having a higher resolution and lower frame rate than the motion capture sequence; compressing the motion image sequence using interframe compression and the accompanying audio and storing the compressed motion image and audio data; and compressing the still images using intraframe coding and storing the compressed still image data.
1. A method for simultaneously recording motion and still images in a digital camera, comprising:
a) using an image sensor to capture a sequence of image frames, the sequence of image frames including a repeating sequence of full resolution image frames regularly interspersed between reduced resolution image frames, wherein the full resolution image frames represent images with more pixels than are represented by the reduced resolution image frames;
b) processing the sequence of image frames using a processor, wherein the processing time for the full resolution image frames is longer than the processing time for the reduced resolution image frames; and
c) storing the processed sequence of image frames.
2. The method of
3. The method of
4. The method of
This application is a continuation of U.S. patent application Ser. No. 09/685,998, filed Oct. 11, 2000, entitled “A DIGITAL CAMERA FOR CAPTURING A SEQUENCE OF FULL AND REDUCED RESOLUTION DIGITAL IMAGES AND STORING MOTION AND STILL DIGITAL IMAGE DATA” by Loui et al., the disclosure of which is incorporated herein by reference.
The invention relates generally to the field of photography, and in particular to combine motion and still image capture. More specifically, the invention relates to a motion/still image capture system which captures image sequences including both full and reduced resolution images and stores digital image data to provide both motion video sequences and digital still pictures.
Currently, it is common for consumers to use camcorders to capture motion video images, which are viewed on a TV display, and film cameras to capture still images which are developed and optically printed to provide hardcopy prints. Carrying both types of devices is burdensome, however. Thus, there is interest in having a digital camcorder type device that captures motion images, while also allowing high quality prints to be produced from selected image frames.
Recent advances in digital camcorders include the ability to capture both motion and still images (commonly referred to as MOST cameras), and associated audio information, such as those from JVC (GR-DV1) and Sony Corp. (DCR-PC7) which allow the capture of motion video and still imagery. For example, the GR-DV1 from JVC allows a user to capture a snapshot while recording live video. Basically, the snapshot is indicated by overlaying a white border on the particular still frame of the captured live video. See U.S. Pat. No. 5,382,974, issued Jan. 17, 1995 to Soeda et al. entitled Movie Camera Having Still Picture Photographing Function and Method of Photographing Still Picture Therewith, which shows a movie camera capable of also capturing still images. Although these cameras allow one to capture motion and still images, they do not provide adequate still image quality. Furthermore, the user must decide before taking the picture whether a still or motion type image is desired. Often, the user would like to have both a motion video sequence and a still image of the same event (e.g. their daughter scoring the winning soccer goal). This is difficult to do with these prior art systems, where the user must press a special button to obtain a “still” image, rather than deciding at a later time what images they would like to have as high resolution prints.
A high definition camcorder could store a high resolution (e.g. 1280×960 pixels) motion sequence for each image frame, to provide higher quality still images. However, to produce good motion rendition of high speed objects, a high frame rate (e.g. 30 frames/sec) is required. The amount of data that would need to be processed and stored per second (e.g. 1280×960×30 pixels) would be excessive, requiring a high cost, power-hungry design having a limited recording time.
There is a need therefore for a digital motion/still camera design that reduces the processing and storage requirements, while providing both good motion rendition and the ability to obtain high resolution prints from any desired scene after it has been captured.
The need is met according to the present invention by providing a method for simultaneously recording motion and still images, that includes the steps of: capturing a motion image sequence and accompanying audio of a scene with a digital video camera, adapted to record both motion and higher resolution still images; simultaneously capturing a still image sequence having a higher resolution and lower frame rate than the motion capture sequence; compressing the motion image sequence using interframe compression and the accompanying audio and storing the compressed motion image and audio data; and compressing the still images using intraframe coding and storing the compressed still image data.
The present invention provides a better way of capturing, processing, and storing both motion and still images of the same event. Another advantage of this invention is that it records both still and motion images so that desired still images can later be selected for printing. Another advantage of this invention is that it uses the same image sensor and digital image processor to provide both the motion and still images. Another advantage of this invention is that the still image processing does not need to be completed within the period set by the capture frame rate, so that the processor pixel operating rate can be reduced by balancing the processor load for still and motion images.
To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures.
Compression of the motion sequence is necessary to lower the bandwidth and storage requirements. For instance, an NTSC format video signal with a frame rate of 29.97 Hz when digitized will result in an uncompressed bit rate of about 168 Mb/s. MPEG (Moving Picture Expert Group) compression of an NTSC video signal can result in a bit rate of 3 to 6 Mb/s with a quality comparable to analog cable TV and far superior to VHS video tape. The full resolution images used to provide high quality still images are obtained much less frequently than the motion images, and have higher resolution and less aggressive compression.
A regular sequence of high resolution still image frames and low resolution compressed audio/video frames is generated 22. Such a sequence is created whenever the user presses the record button while the camera 12 is in the combination motion/still (MOST) mode. The camera 12 takes a fixed number of images per second (e.g. 24 or 30 frames per second) to provide the motion sequence. A relatively small number of these images (e.g. 2 frames per second, or every 12th or 15th image) are normally full resolution images, while the others are reduced resolution images. The user can set the frequency of full resolution images using user controls 66, in order to record more high resolution images per second for important but short events, or alternatively fewer high resolution images per second, to reduce the amount of still image related data that must be recorded, in order to increase the recording time.
The full and reduced resolution images can be created using the image sensor described in commonly-assigned U.S. Pat. No. 5,440,343 issued Aug. 8, 1995 to Parulski et al., entitled Motion/Still Electronic Image Sensing Apparatus, the disclosure of which is incorporated herein. Alternatively, the image sensor can always read out a full resolution image using a relatively high speed clock, and the reduced resolution images can be obtained by averaging or subsampling (e.g. decimating the image 2:1 both vertically and horizontally) this full resolution image sequence.
The low resolution video data and high resolution image data can be stored 24 on a digital storage medium, such as CD ROM medium or a DVD (Digital Versatile Disk) medium 26. Alternatively, the data can be stored on digital videotape, a magnetic hard drive, or Flash EPROM memory, or any other digital storage device. Alternatively, the data can be stored on an image server through a wireless or wired network link. The stored data may also include metadata, such as user-input text information and camera-generated information such as time, date, camera model and serial number, lens settings, etc.
The structure of the image files that store the low resolution motion video and high resolution still image data are described below. The display and printing of these digital image files 28 are carried out at a user's video player terminal (which may utilize the digital camera 12 for image playback) or computer, and are enabled by an image processing sub-system in the terminal or computer that decompresses the image files. Through the terminal or computer, the user can view and browse the low resolution video sequences, and selectively print any associated high resolution still frames.
A user interface of the terminal or computer is depicted in
The architecture of the digital motion/still camera 12 is illustrated in
The digital data from ASP and A/D 52 includes both full resolution image frames that are used to produce both high resolution still image data and a low resolution motion image data, and reduced resolution image frames that are only used to provide low resolution motion image data. As described earlier, the image sensor 50 can use the invention described in U.S. Pat. No. 5,440,343 to provide full resolution (e.g. 1280×960 pixel) images and reduced resolution (e.g. 640×480 pixel) images by summing and dumping appropriate color pixels. Alternatively, the sensor can provide only full resolution (e.g. 1280×960 pixel) images, and a 2:1 subsampling circuit 54 can be used to produce the reduced resolution image by appropriately averaging the digital sensor pixel values. The ASP and A/D 52 is coupled to a full resolution buffer memory 60 which stores a full resolution (e.g. 1280×960 pixel) still image. The 2:1 subsampling circuit 54 is coupled to a multi-frame reduced resolution buffer memory 62, which can store multiple reduced resolution images, such as buffer memory that can store four 640×480 pixel images.
The outputs of the full resolution buffer memory 60 and the multi-frame reduced resolution buffer memory 62 are coupled to a multiplexer switch 64 which provides an input to control microprocessor 68. The digital audio data from audio A/D 58 is also input to control microprocessor 68. The processed image and audio data is coupled to a digital recorder 48 which stores the digital image files. The digital recorder can use write-once or erasable CD or DVD optical disks. Alternatively, the digital recorder can use other digital storage technologies, such as magnetic hard drives, magnetic tape, optical tape, or solid-state memory.
The control microprocessor 68 alternates between processing one full resolution image from full resolution buffer memory 60, followed by several reduced resolution images from multi-frame reduced resolution buffer memory 62. As will be described later in reference to
For each full and reduced resolution image frame, the control microprocessor 68 performs color interpolation followed by color and tone correction, in order to produce rendered sRGB image data. For full resolution frames, the control microprocessor 68 creates both full resolution (e.g. 1280×960 pixel) sRGB image data, and low resolution (e.g. 640×480 pixel) sRGB image data. For reduced resolution frames, the digital processor creates only low resolution (e.g. 640×480 pixel) sRGB data. The control microprocessor 68 also creates, from the low resolution image data of each full and reduced resolution image frame, a “thumbnail” size image, which can be created as described in commonly-assigned U.S. Pat. No. 5,164,831 issued Nov. 17, 1992 to Kuchta et al., entitled Electronic Still Camera Providing Multi-Format Storage of Full and Reduced Resolution Images, the disclosure of which is herein incorporated by reference. This thumbnail image has 160×120 pixels and is displayed on electronic viewfinder 72 as the image sequences are captured. The digital processor 66 can also provide a video output signal 74, as either a digital or analog video signal, which is encoded for display on a standard TV display (not shown), such as an NTSC video display. The digital processor also compresses the digital audio signal from audio A/D 52, in order to provide an MPEG compressed audio stream.
The rendered sRGB low resolution image data from both the full and reduced resolution frames is then compressed, processed and stored, along with the compressed audio data, as an MPEG-2 bitstream. The processing may include conversion form RGB data to alternate color spaces, such as YUV, YIQ, or Y, Cb, Cr as described in chapter 3 of “Video Demystified” by Keith Jack, published by HighText Interactive Inc., SanDiego, Calif. copyright 1996. In the MPEG-2 standard, frames are designated I, P or B. I indicating that the frame is intra-coded (the encoding is not dependent on any other frame); P indicating the frame is predicted from the previous frame; and B indicating that the frame is predicted from both the previous and future frames. In one embodiment of the present invention, the low resolution image frames derived by subsampling the full resolution image frames are preferably compressed as I frames. This is because each I frame can be decoded as a low resolution image, so that the full resolution image data can be stored as the difference between this I frame low resolution image and the full resolution image.
The rendered sRGB full resolution image data from the full resolution frames is also compressed and stored to provide high resolution still image data. In one embodiment, the difference (residual) between the sRGB full resolution image data, and the upsampled low resolution I frame image decompressed from the MPEG-2 bitstream is computed, and then compressed and stored as a complete still image file. To recover the full resolution still image data, the residual still image file and the corresponding MPEG-2 I frame are both decompressed and combined. In an second embodiment, the full resolution still image data is JPEG compressed and stored as a complete still image file. The MPEG-2 and still image files can store metadata, such as the date and time the picture was captured, the lens f/number and other camera settings, and image captions or comments that can be selected or entered by the camera user.
The motion capture and processing sequence used in the digital camera 12 of the present invention will now be described with reference to
In conventional camcorders, the processing time for each video frame is equal to one frame period. In the present invention, however, the processing time for the full resolution frames is set to be substantially longer than the processing time for the reduced resolution frames. As shown in
As shown in
The MPEG-2 frames corresponding to the reduced resolution frames 104 (see
The invention has been described with reference to a preferred embodiment. However, it will be appreciated that variations and modifications can be effected by a person of ordinary skill in the art without departing from the scope of the invention.