TECHNICAL FIELD

[0001] The present disclosure relates to the field of digital video analysis and encoding, particularly a method of detecting long shots in sports video.

BACKGROUND

[0002] Watching sports video is a popular pastime for many people. Digital transmissions of sports video can be viewed on televisions directly or through set-top boxes, or on other devices such as personal computers, tablet computers, smartphones, mobile devices, gaming consoles, and/or other equipment. Digital recordings of sports video can be viewed on the same devices and viewing the digital recordings can begin at the start of a recorded event or midway through the event.

[0003] Automatic parsing of sports video based on visual and/or cinematographic cues can be used to identify segments of potential interest to a viewer, and/or points at which video on demand playback can begin. Visual cues, such as long shots, medium shots and close up shots, can be used to identify segments of the video where on-field events are occurring, or to distinguish on-field events from close up views of players, referees, balls, logos, or other items. Long shots frequently provide coverage of large areas of a playing surface, such as a playing field or court, and frequently identify periods of time during which activity on the field is at a maximum. Extended periods of play can comprise a sequence of long shots followed by medium and/or close up shots which signify the end of a play or highlight the contributions of key players. Detection of long shots can also aid in automatically identifying highlights from the video, and/or automatically summarizing video.

[0004] Some methods of detecting long shots have been developed. However, most existing methods require training time and cannot immediately identify long shots based on a single frame of video. For example, some methods create hue histograms from randomly selected frames over a period of time, and then identify peaks and thresholds from the histogram. These methods require time for their model to be trained at the beginning of a sports video, and the training can be impeded when non-sports video is interspersed with the sports video, such as commercials or pre-game analysis, or if portions of the field are covered with statistics, logos or other graphics. In other methods, the color of a playing surface is learned at the beginning of a sports video, but the method can need time to learn the color of the playing surface before the long shot detection algorithm can begin. In still other methods, the color of the playing surface must be preidentified, and/or compares the color of the playing surface to a reference model.

SUMMARY

[0005] What is needed is a method for detecting long shots in sports video that can operate on a single frame without needing analysis of multiple frames over a period of time, or needing prior information about the color of a playing surface. The present disclosure provides a method for a device receiving a video frame to determine whether the frame is a long shot by dividing the video frame into candidate field pixels and candidate non-field pixels, determining whether enough pixels are candidate field pixels, determining whether the hue of those candidate field pixels is within a typical range, and whether one or more standard deviations are small enough to confirm that the candidate field pixels have a consistent hue. [0006] In one embodiment, the present disclosure includes a method of identifying long shots of sports video is provided, the method comprising receiving a video frame comprising a plurality of pixels, classifying each of the plurality of pixels as a candidate field pixel or a candidate non-field pixel, determining whether at least a predefined percentage of the plurality of pixels are candidate field pixels, calculating a first standard deviation, the first standard deviation being the standard deviation of the hues of all candidate field pixels, and classifying the video frame as a long shot of sports video when at least the predefined percentage of the plurality of pixels are candidate field pixels and the first standard deviation is equal to or lower than a predefined maximum standard deviation value.

[0007] In another embodiment, the present invention includes a system, the system comprising a memory configured to receive and store a video frame comprising a plurality of pixels, and a processor configured to classify each of the plurality of pixels as a candidate field pixel or a candidate non-field pixel, determine whether at least a predefined percentage of the plurality of pixels are candidate field pixels, calculate a first standard deviation, the first standard deviation being the standard deviation of the hues of all candidate field pixels, classify the video frame as a long shot of sports video when at least the predefined percentage of the plurality of pixels are candidate field pixels and the first standard deviation is equal to or lower than a predefined maximum standard deviation value.

BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Further details of the present invention are explained with the help of the attached drawings in which:

[0009] Fig . 1 depicts an exemplary frame of a long shot of sports video.

[0010] Fig . 2 depicts a method for determining whether a frame is a long shot of sports video.

[0011] Fig . 3 depicts exemplary code that can implement steps of the method shown in Fig. .

[0012] Fig . 4A depicts an exemplary frame of a long shot of sports video.

[0013] Fig . 4B depicts a histogram of the hues of the frame shown in Fig. 4A.

[0014] Fig . 5 A depicts an exemplary frame of a long shot of sports video.

[0015] Fig . 5B depicts a histogram of the hues of the frame shown in Fig. 5A.

[0016] Fig . 6A depicts an exemplary frame of a long shot of sports video.

[0017] Fig . 6B depicts a histogram of the hues of the frame shown in Fig. 6A.

[0018] Fig . 7A depicts an exemplary frame of a long shot of sports video.

[0019] Fig . 7B depicts a histogram of the hues of the frame shown in Fig. 7A.

[0020] Fig . 8 depicts an exemplary embodiment of computer hardware that can perform the disclosed embodiments and methods.

DETAILED DESCRIPTION

[0021] Fig. 1 depicts an exemplary frame 100 of a long shot in sports video. A video frame 100 can comprise a plurality of pixels. By way of a non-limiting example, in some 1080p high definition videos, each frame 100 can comprise 1920 pixels by 1080 pixels. Long shots can be shots of one or more subjects that are taken at a distance and/or have a wide field of view. In sports video, long shots generally show a portion of a playing surface 102, as shown in Fig. 1. In some sports, the playing surface 102 can be a field, such as in football, baseball, and soccer. In other sports, the playing surface 102 can be a court, such as in basketball or tennis. In still other sports, the playing surface 102 can be an ice rink, such as in ice hockey, or other type of playing surface. Playing surfaces 102 can be made of grass, turf, wood, ice, cement, clay, sand, any other type of surface. Regardless of the specific sport, most playing surfaces 102 have large areas of uniform or near uniform color. By way of non-limiting examples, in football most fields are made of grass or turf that is predominantly green, while in basketball many courts are made of hardwood that is predominantly tan. Moreover, in some embodiments and/or situations, large playing surfaces 102 can be more or less bright than surrounding areas. By way of a non- limiting example, the court surface in a basketball video can be illuminated with much higher intensity lights than the areas immediately adjacent to the court.

[0022] As can be seen from Fig. 1, in long shots the playing surface 102 can be at least partially obscured by offensive and/or defensive players 104, referees, balls, lines 106, logos, onscreen displays 108, or other items. In some long shots other elements or backgrounds can be visible, such as fans, stands, sidelines, portions of a stadium, or other items. Even though other items can be visible in the frame 100, in most long shots the uniform or near uniform color of the playing surface 102 can still fill a significant portion of the frame, as shown in Fig. 1. When long shots can be identified in sports video, this information can be used to influence the operations of video processing equipment that includes video encoders and/or transcoders. For example, in some embodiments a video encoder can modify its operating parameters such as quantization step size, coding unit size, transform unit size and macro-block modes based on the presence of a long shot being signaled. In some embodiments, video encoding can be optimized because most playing surfaces 1 10 have a uniform or near uniform color, and therefore have less detail than other areas of the image such as the players, fans, or backgrounds other than the playing surface 102. Additionally, video frames classified as long shots frequently contain areas of considerable interest, such as the play field and active participants in the events, adjacent to areas of much less interest to most viewers, such as fans in the stands.

[0023] In some embodiments, a video preprocessing system communicatively coupled with the input of a video encoder can perform media analysis and content classification on behalf of the encoder, allowing the encoder to devote less time and fewer bits to encoding the portions of the frame 100 devoted to the uniform or near uniform playing surface 102 and regions outside the playing surface 102, and more encoding time and effort can be devoted to areas of the image that have more detail and relevance to a viewer.

[0024] Fig. 2 depicts a method for determining long shots in sports video. The method of Fig. 2 can find that a frame 100 is a long shot when the hues of its pixels have a bimodal distribution biased toward hues that are typical for the playing surface 102 of the sporting event, and also have a tight standard deviation around a hue typical for the playing surface 102.

[0025] At step 200, a media analysis system can receive a video frame 100 comprising a plurality of pixels. The system can be a video preprocessing system, video processor, encoder, server, computer, or any other piece of equipment that can receive and analyze video frames 100. [0026] The video frame 100 can be received from a source, such as a video decoder, camera, media server, receiver, or any other type of television or video transmission equipment capable of transporting uncompressed and/or compressed video signals. By way of a non-limiting example, in some embodiments the video frame 100 can be received from a video decoder that takes a compressed video signal received on a local area network or from the output of a satellite, terrestrial, or cable television receiver and expands it into an uncompressed component video signal, such as a RGB (Red, Green, Blue) signal.

[0027] At step 202, the system can determine the HSV (Hue, Saturation, and Value) components of each pixel in the frame 100. The HSV components of a pixel can together describe the color of the pixel. The hue component can describe the pixel's hue or tint. The saturation component can describe the pixel's saturation or amount of gray. The value component can describe the pixel's brightness or color intensity. In some embodiments, the color of the pixels can initially be described with the RGB (Red, Green, Blue) color model, and the RGB values can be converted into HSV values.

[0028] At step 204, each pixel of the frame can be classified as a candidate field pixel 1 10 or a candidate non-field pixel 1 12. In some embodiments, a hue component threshold 1 14 can be determined, and the system can classify all pixels having a hue component below the hue component threshold 1 14 as a candidate field pixel 1 10, and all pixels having a hue component above the hue component threshold 1 14 as a candidate non-field pixel 1 12. Division of the pixels at a hue component threshold 100 can be appropriate as a step to determine long shots because the hues of the frames 100 of long shots generally have bimodal distributions, while the hues of the frames 100 of close up shots and/or medium shots generally do not have bimodal distributions. [0029] In some embodiments, the hue component threshold 1 14 can be determined using an image thresholding method. In some embodiments, a histogram-based image thresholding method can be used, in which a histogram 1 16 of the hue components of all pixels in the frame 100 is generated and the histogram 1 16 is divided into two sections, the candidate field pixels 1 10 and the candidate non-field pixels 1 12, at the hue component threshold 1 14. By way of a non-limiting example, in some embodiments the hue component threshold 1 14 can be determined using Otsu's method, in which the hue component threshold 1 14 can be determined to be the hue level on the histogram 1 16 at which the combined spread of the hue components of the pixels on either side of the hue component threshold 1 14 is at its minimum. In alternate embodiments, the hue component threshold 1 14 can be determined using clustering-based methods, entropy-based methods, object attribute -based methods, spatial methods, local methods, and/or any other desired image thresholding method.

[0030] In alternate embodiments, a value component threshold can be determined in addition to the hue component threshold 1 14 and the system can classify all pixels having a hue component below the hue component threshold 1 14 and above the value component threshold as a candidate field pixel 1 10, and all pixels having either a hue component above the hue component threshold 1 14 or a value component below the value component threshold as a candidate non-field pixel 1 12. Division of the pixels at the value component threshold can be appropriate as a step to determine long shots because the values of the frames 100 of long shots often have bimodal distributions in both the hue and the value components, for example when the playing surface 102 is illuminated by stadium lights to a greater extent than surrounding areas, while the hue and value components of the frames 100 of close up shots and/or medium shots generally do not have highly correlated bimodal distributions. [0031] In some embodiments, the value component threshold can be determined using an image thresholding method similar to the determination of the hue component threshold 1 14. In some embodiments, a histogram-based image thresholding method can be used, in which a histogram 1 16 of the value components of all pixels in the frame 100 is generated and the histogram 1 16 is divided into two sections, the candidate field pixels 1 10 and the candidate non- field pixels 1 12, at the value component threshold. By way of a non-limiting example, in some embodiments the value component threshold can be determined using Otsu's method, in which the value component threshold can be determined to be the value level on the histogram 1 16 at which the combined spread of the value components of the pixels on either side of the value component threshold is at its minimum. In alternate embodiments, the value component threshold can be determined using clustering-based methods, entropy-based methods, object attribute -based methods, spatial methods, local methods, and/or any other desired image thresholding method.

[0032] The classification of pixels as candidate field pixels 1 10 and candidate non-field pixels 1 12 based on the hue component threshold 1 14 of a single frame 100 or both the hue component threshold 1 14 and the value component threshold of a single frame can allow the method of Fig. 2 to function regardless of the specific hues and/or values on the histogram 1 16, such that the method can be used with any type or color of playing surface 102, and any lighting or weather conditions. Because the method classifies the pixels of a single frame 100, the method does not require training time over multiple frames 100 and is not affected by changing lighting or weather conditions over multiple frames 100.

[0033] At step 206, the system can determine whether at least a predefined percentage of the pixels in a frame 100 are candidate field pixels 1 10. In some embodiments, the predefined percentage can be 50%, such that the system can determine whether the candidate field pixels 1 10 take up a majority of the frame 100. In alternate embodiments, the predefined percentage can be any desired percentage.

[0034] In some embodiments when the predefined percentage is 50%, the system at step 206 can determine whether the majority of the pixels in the frame 100 are candidate field pixels 1 10 by calculating a total mean hue 1 18, the mean hue of all the pixels in the frame 100. The system can compare the total mean hue 1 18 to the hue component threshold 1 14 determined in step 204. If the total mean hue 1 18 is above the hue component threshold 1 14, the system can determine that the majority of the pixels in the frame 100 are candidate field pixels 1 10.

[0035] If the system determines at step 206 that less than the predefined percentage of the pixels in the frame 100 are candidate field pixels 1 10, the system can determine that the frame 100 is not a long shot, and can move to step 220 to inform an encoder that the frame 100 is not a long shot and/or inform the encoder to encode the frame normally without considering the frame 100 to be a long shot. In some embodiments, the encoder can be a part of the system itself. In alternate embodiments, the encoder can be a separate component in communication with the system. If the system determines at step 206 that at least the predefined percentage of the pixels in the frame 100 are candidate field pixels 1 10, the system can move to step 208.

[0036] In some embodiments, at step 208 the system can calculate the field mean hue 120, the mean hue of the pixels in the frame 100 that were determined to be candidate field pixels 1 10 in step 204. The mean field hue 120 can be obtained by averaging the hue components of the candidate field pixels 1 10.

[0037] In some embodiments, at step 210 the system can determine whether the field mean hue 120 of the candidate field pixels 1 10 is within a predefined range of values. In some embodiments, the predefined range of values can be a range of values for the hue component that are typical for the playing surface 102 of the particular sporting event and/or venue being shown in the video. The playing surfaces 102 of different sports and/or specific venues can have different typical ranges of hue components. By way of a non-limiting example, a grass football field can have a typical range of hue components describing green hues.

[0038] In alternate embodiments, steps 208 and 210 can be absent, such that the field mean hue 120 of the candidate field pixels 1 10 is not determined and is not compared to a predefined range of values. In these alternate embodiments, the system can move directly to steps 212 and/or 214 if the system determined during step 206 that at least the predefined percentage of the pixels in a frame 100 were candidate field pixels 1 10.

[0039] If the system determines at step 210 that the field mean hue 120 is not within the predefined range of values, the system can determine that the frame 100 is not a long shot, and can move to step 220 to inform the encoder that the frame 100 is not a long shot and/or inform the encoder to encode the frame normally without considering the frame 100 to be a long shot. . If the system determines at step 210 that the field mean hue 120 is within the predefined range of values, the system can move to steps 212 and/or 214.

[0040] At step 212, the system can calculate a first standard deviation 122, the standard deviation of the hue components of all of the candidate field pixels 1 10 in the frame 100. In some embodiments, at step 214, the system can also calculate a second standard deviation 124, the standard deviation of the hue components of the subset of candidate field pixels 1 10 that have hue components that fall within a predefined range of values. As discussed above, in some embodiments the predefined range of values can be the range of values for the hue component that are typical for the playing surface 102 of the particular sporting event and/or venue being shown in the video. In alternate embodiments, step 214 can be absent, such that only the first standard deviation 122 is calculated at step 212.

[0041] In some embodiments, the field mean hue 120, the first standard deviation 122 and/or the second standard deviation 124 can be calculated directly from the histogram 1 16. The field mean hue 120 can be calculated from hue histogram values that are less than the hue component threshold 1 14. By way of a non-limiting example, the field mean hue 120 can equal a first histogram sum divided by a second histogram sum, such that the first histogram sum equals the sum of the product of each hue value that is less than the hue component threshold 1 14 multiplied by its histogram value, and such that the second histogram sum equals the sum of histogram values that have a hue less than the hue component threshold 1 14. Similarly, a field mean square hue can be calculated by dividing a third histogram sum by the second histogram sum, such that the third histogram sum equals the sum of the product of each hue value that is less than the hue component threshold 1 14 multiplied by its histogram value squared. The first standard deviation 122 can be calculated from the field mean hue and the field mean square hue by taking the square root of the sum of the field mean square minus the field mean hue squared. The second standard deviation 124 can be calculated from a predefined range mean hue and a predefined mean square hue using the same technique used for the first standard deviation 122 calculation, except that the predefined range mean hue and the predefined range mean square hue are computed from histogram values within the bounds of a predefined range instead of from histogram values less than the hue component threshold 1 14. As discussed above, the predefined range of values can be a range of values for the hue component that are typical for the playing surface 102 of the particular sporting event and/or venue being shown in the video. In some situations and/or embodiments in which the number of candidate field pixels 1 10 is larger than the number of hue values, calculating the mean hue and standard deviation values from the hue histogram 1 16 can be computationally more efficient than calculating these values directly than from candidate field pixels 1 10.

[0042] In some embodiments at step 216, the system can determine whether at least one of the first standard deviation 122 and the second standard deviation 124 that were calculated in steps 212 and 214 are equal to or lower than a predefined maximum standard deviation value. In some embodiments, the predefined maximum standard deviation value can be selected such that when at least one of the first standard deviation 122 and the second standard deviation 124 is lower than the predefined maximum standard deviation value, the system can have detected candidate field pixels 1 10 that represent a playing surface 102 that are unified by a dominant overall hue. In some situations and/or embodiments, the playing surface 102 can have secondary playfield features that have one or more colors that are different than the dominant overall hue, such that the different colors can stand out to television and live audiences and/or designate different portions of the playing surface 102 for players and game officials. By way of non- limiting examples, in some basketball courts the key areas underneath the basket can be a different color than the rest of the court, and in some football fields the end zone area can be a different color than the rest of the field. In some embodiments, the second standard deviation 124, which is calculated over a limited range of hues, can be used to isolate the dominant overall hue of the playing surface 102 from secondary playfield features that have different colors.

[0043] In alternate embodiments, at step 216 the system can determine whether the first standard deviation 122 that was calculated in step 212 is equal to or lower than the predefined maximum standard deviation value without considering the second standard deviation 124, for example in embodiments in which step 214 is absent. [0044] If the system determines at step 216 that neither of the first standard deviation 122 and the second standard deviation 124 is equal to or lower than the predefined maximum standard deviation value, the system can determine that the frame 100 is not a long shot, and can move to step 220 to inform the encoder that the frame 100 is not a long shot and/or inform the encoder to encode the frame normally without considering the frame 100 to be a long shot. If the system determines at step 210 that one or both of the first standard deviation 122 and the second standard deviation 124 is equal to or lower than the predefined maximum standard deviation value, the system can determine that the frame 100 is a long shot and can move to step 218 to inform the encoder that the frame 100 is a long shot. In some embodiments, at step 218 the system can also inform the encoder which pixels in the long shot frame 100 were determined to be candidate field pixels 1 10.

[0045] After the system has informed the encoder that the frame 100 is a long shot at 218, or has informed the encoder that the frame 100 is not a long shot or should be encoded with normal encoding without considering the frame 100 to be a long shot at step 220, the system can return to step 200 to receive the next frame 100 in the video from the source.Fig. 3 depicts a non- limiting exemplary embodiment of Python code that can be used to implement to process depicted in Fig. 2. In Fig. 3, step 210 can be implemented using the variables "bin mean otsu," which can be the mean field hue 120, and "lowerLongShotThresh" and "upperLongShotThresh," which can be the lower and upper bounds of a predefined range of values, for example the range of values for the hue component typical for the playing surface 102 of the particular sporting event and/or specific venue being shown in the video. Step 216 can be implemented in Fig. 3 using the variables "bin_std_otsu," which can be the first standard deviation 122,

"bin_std_range," which can be the second standard deviation 124, and "stdHueThresh," which can be the predefined maximum standard deviation value. Step 206 can be implemented in Fig. 3 using the variables "bin mean hue," which can be the total mean hue 1 18, and "int(t hue)," which can be the hue component threshold 1 14. Step 218, the step of informing the encoder that the frame 100 is a long shot, can be indicated by setting "longShot" to "true." It should be noted that in the embodiments of the method shown in Fig. 2, Fig. 3, or any other embodiment, the tests of steps 206, 210, and/or 216 can be performed in any order.

[0046] Figs. 4A-7B depict examples of video frames 100 and their associated histograms 1 16. Each exemplary histogram 1 16 depicts the frame's hue component threshold 1 14, total mean hue 1 18, field mean hue 120, first standard deviation 122, and second standard deviation 124.

[0047] Fig. 4A depicts an exemplary frame 100 showing a long shot of a football game taken from a high angle. The method of Fig. 2 can determine that the frame 100 shown in Fig. 4A is a long shot of sports video. As can be seen from the histogram of Fig. 4B, the system at step 204 can determine the hue component threshold 1 10 and divide the pixels having hue components greater than the hue component threshold 1 10 as candidate non-field pixels and the pixels having hue components less than the hue component threshold 1 10 as candidate field pixels. The system at step 206 can determine that a majority of the pixels are candidate field pixels because the total mean hue 1 12 is less than the hue component threshold 1 10. The system at step 210 can find that the field mean hue 1 14 is within the typical range for the sporting event. The system at step 216 can find that both the first standard deviation 1 16 and the second standard deviation are less than the predefined maximum standard deviation value. Because the frame 100 passes the mean tests of steps 206 and 210, and the standard deviation test of step 216, the frame 100 of Fig. 4A can be determined to be a long shot of sports video. [0048] Fig. 5A depicts an exemplary frame 100 showing a long shot of vehicles in front of a house, near a green lawn. Although the green lawn has many of the same hues as a green football field, the method of Fig. 2 can determine that the frame 100 shown in Fig. 5A is not a long shot of sports video. As can be seen from the histogram of Fig. 5B, the system at step 204 can determine the hue component threshold 1 10 and divide the pixels having hue components greater than the hue component threshold 1 10 as candidate non-field pixels and the pixels having hue components less than the hue component threshold 1 10 as candidate field pixels. The system at step 206 can determine that a majority of the pixels are candidate field pixels because the total mean hue 1 12 is less than the hue component threshold 1 10. The system at step 210 can find that the field mean hue 1 14 is within the typical range for the sporting event, even though that sporting event is not shown in the frame 100. However, the system at step 216 can find that neither of the first standard deviation 1 16 and the second standard deviation are less than the predefined maximum standard deviation value. Although the frame 100 passes the mean tests of steps 206 and 210, the frame 100 fails the standard deviation test of step 216 and therefore the frame 100 of Fig. 5A can be determined to not be a long shot of sports video.

[0049] Fig. 6A depicts an exemplary frame 100 showing a close-up shot of a football player. Although the frame 100 shows the sport of football, the method of Fig. 2 can determine that the frame 100 shown in Fig. 6A is not a long shot of sports video. As can be seen from the histogram of Fig. 6B, the system at step 204 can determine the hue component threshold 1 10 and divide the pixels having hue components greater than the hue component threshold 1 10 as candidate non- field pixels and the pixels having hue components less than the hue component threshold 1 10 as candidate field pixels. The system at step 206 can determine that less than a majority of the pixels are candidate field pixels because the total mean hue 1 12 is greater than the hue component threshold 1 10. The system could inform the encoder that the frame 100 has been determined to not be a long shot at this point, however if the process were to continue the system at step 210 can also find that the field mean hue 1 14 is not within the typical range for the sporting event. The system could also inform the encoder that the frame 100 has been determined to not be a long shot at this point. However, if the process were to continue even further, the system at step 216 can also find that neither of the first standard deviation 1 16 and the second standard deviation are less than the predefined maximum standard deviation value. Because the frame 100 fails the mean tests of steps 206 and 210, and also fails the standard deviation test of step 216, the frame 100 of Fig. 6A can be determined to not be a long shot of sports video.

[0050] Fig. 7A depicts an exemplary frame 100 showing a medium shot of travelers waiting at an airport. The method of Fig. 2 can determine that the frame 100 shown in Fig. 7A is not a long shot of sports video. As can be seen from the histogram of Fig. 7B, the system at step 204 can determine the hue component threshold 1 10 and divide the pixels having hue components greater than the hue component threshold 1 10 as candidate non-field pixels and the pixels having hue components less than the hue component threshold 1 10 as candidate field pixels. The system at step 206 can determine that a majority of the pixels are candidate field pixels because the total mean hue 1 12 is less than the hue component threshold 1 10. However, the system at step 210 can find that the field mean hue 1 14 is not within the typical range for the sporting event. The system could inform the encoder that the frame 100 has been determined to not be a long shot at this point, however if the process were to continue the system at step 216 can also find that neither of the first standard deviation 1 16 and the second standard deviation are less than the predefined maximum standard deviation value. Although the frame 100 passes the mean test of steps 206, the frame 100 fails the mean test of step 210 and also fails the standard deviation test of step 216 and therefore the frame 100 of Fig. 7 A can be determined to not be a long shot of sports video.

[0001] The execution of the sequences of instructions required to practice the embodiments may be performed by a computer system 800 as shown in Fig. 8. In an embodiment, execution of the sequences of instructions is performed by a single computer system 800. According to other embodiments, two or more computer systems 800 coupled by a communication link 815 may perform the sequence of instructions in coordination with one another. Although a description of only one computer system 800 may be presented herein, it should be understood that any number of computer systems 800 may be employed.

[0002] A computer system 800 according to an embodiment will now be described with reference to Fig. 8, which is a block diagram of the functional components of a computer system 800. As used herein, the term computer system 800 is broadly used to describe any computing device that can store and independently run one or more programs.

[0003] The computer system 800 may include a communication interface 814 coupled to the bus 806. The communication interface 814 provides two-way communication between computer systems 800. The communication interface 814 of a respective computer system 800 transmits and receives electrical, electromagnetic or optical signals, that include data streams representing various types of signal information, e.g., instructions, messages and data. A communication link 815 links one computer system 800 with another computer system 800. For example, the communication link 815 may be a LAN, an integrated services digital network (ISDN) card, a modem, or the Internet. [0004] A computer system 800 may transmit and receive messages, data, and instructions, including programs, i.e., application, code, through its respective communication link 815 and communication interface 814. Received program code may be executed by the respective processor(s) 807 as it is received, and/or stored in the storage device 810, or other associated non-volatile media, for later execution.

[0005] In an embodiment, the computer system 800 operates in conjunction with a data storage system 831 , e.g., a data storage system 831 that contains a database 832 that is readily accessible by the computer system 800. The computer system 800 communicates with the data storage system 831 through a data interface 833.

[0006] Computer system 800 can include a bus 806 or other communication mechanism for communicating the instructions, messages and data, collectively, information, and one or more processors 807 coupled with the bus 806 for processing information. Computer system 800 also includes a main memory 808, such as a random access memory (RAM) or other dynamic storage device, coupled to the bus 806 for storing dynamic data and instructions to be executed by the processor(s) 807. The computer system 800 may further include a read only memory (ROM) 809 or other static storage device coupled to the bus 806 for storing static data and instructions for the processor(s) 807. A storage device 810, such as a magnetic disk or optical disk, may also be provided and coupled to the bus 806 for storing data and instructions for the processor(s) 807.

[0007] A computer system 800 may be coupled via the bus 806 to a display device 81 1 , such as an LCD screen. An input device 812, e.g., alphanumeric and other keys, is coupled to the bus 806 for communicating information and command selections to the processor(s) 807.

[0008] According to one embodiment, an individual computer system 800 performs specific operations by their respective processor(s) 807 executing one or more sequences of one or more instructions contained in the main memory 808. Such instructions may be read into the main memory 808 from another computer-usable medium, such as the ROM 809 or the storage device 810. Execution of the sequences of instructions contained in the main memory 808 causes the processor(s) 807 to perform the processes described herein. In alternative embodiments, hardwired circuitry may be used in place of or in combination with software instructions. Thus, embodiments are not limited to any specific combination of hardware circuitry and/or software.

[0051] Although the present invention has been described above with particularity, this was merely to teach one of ordinary skill in the art how to make and use the invention. Many additional modifications will fall within the scope of the invention, as that scope is defined by the following claims.