Image comparison evidence can be a very powerful form of evidence in court. This kind of expert forensic video analysis evidence is frequently challenged by opposing attorneys. Through this evidence, using the objective lens of the camera and the analyst’s trained eye and detailed analysis, the trier of fact may be better positioned to ascertain what really happened. After all, the ultimate goal of all court proceedings, whether criminal or civil in nature, should be to ascertain the truth.
Image comparison involves the detailed and methodical examination of questioned images followed by known images and then comparing and contrasting the two images, typically for identification purposes. The goal of the comparison process is typically to determine whether the person, vehicle or object in question is in fact the known person, vehicle or object. This type of evidence has been frequently led across North America over the past several years, almost always in criminal cases. It has also been heavily challenged in many cases.
This robust comparison process cannot be properly undertaken by an untrained person. It requires education, training, experience and skill and ultimately calls for an opinion and therefore this process should only be employed by a properly trained expert. There are specialty courses that are taught on this specific subject and anyone wishing to engage in image comparison would be well advised to undertake such training.
The comparison component of image analysis evidence can be of tremendous assistance to the court in ascertaining whether there is a connection between the questioned images and the known images. Comparison evidence seeks to maximize the evidential value of the images. I have led image comparison evidence in court many times over the years and the most common challenge to this type of evidence is that the trier of fact does not require expert assistance in order to form a correct opinion. This challenge is common across North America. It is often argued by opposing counsel that the trier of fact can look at the images and do their own comparison and that comparison evidence by an expert analyst is neither required, nor helpful. This argument misses the mark. By the time the analyst enters the witness stand, he/she will have spent many hours examining, clarifying and interpreting the images and even longer comparing the questioned and known images. The analyst will have written a comprehensive report, often the subject of peer review. The comparison evidence of a skilled analyst who has studied the images far more than the trier of fact ever would and who understands the technical aspects of video evidence is clearly of considerable value in the fact-finding process. It is perhaps for this reason that it is challenged.
Image comparison evidence has been the subject of many court challenges over the years. Courts in the United States have ruled favorably on the use of image comparison evidence in identifying vehicles [i], clothing and other objects [ii], and people [iii].
In State of Michigan v. Fomby, 2013 WL 1137075, 300 Mich.App. 46, 831 N.W.2d. 887 (Court of Appeals of Michigan), a certified forensic video technician testified for the State regarding his examination and comparison of digital images. The purpose of the comparison was to determine if the suspects involved in the shooting that was captured on surveillance video had earlier been to a gas station, such visit also having been recorded on surveillance video. The technician opined that the same people were seen in both locations but did not offer any opinion as to whether the defendant was one of the suspects.
On the appeal of his conviction, the defendant argued that this evidence was lay opinion testimony (rather than expert evidence) and that it was irrelevant and superfluous in any event because such opinions could have been drawn by the jury without assistance. He argued that the evidence invaded the province of the jury and should not have been allowed in evidence at trial.
The Court of Appeals stated at pages 2-4:
The gateway question is whether Gibson’s testimony constituted expert testimony or lay opinion testimony, i.e., whether MRE 702 or MRE 701 applies respectively. MRE 701 is virtually identical to FRE 701, and because no published Michigan case addresses this specific issue, we review relevant federal cases. In United States v. Begay, 42 F3d 486, 502 (CA 9th, 1994), an officer provided narrative testimony regarding an enhanced video of a demonstration involving about 200 demonstrators that resulted in violence. The officer magnified the videotape, reviewed over 800 photographs taken during the incident, copied portions of the videotape in slow motion and enhanced its quality to help his identification of the individuals depicted. He then added circles and arrows to help the jury follow the defendants’ movements. The Ninth Circuit determined that this was not expert testimony; it was lay witness opinion testimony. Just as the officer in Begay presented lay opinion testimony, Gibson also presented lay opinion testimony. Further, Gibson was qualified as a forensic video technician, “proficient in the acquisition, production[,] and presentation of … video evidence in court[.]” Even if these qualifications do not extend to comparison and identification of individuals within still photographs made from videos, Gibson’s testimony was properly admitted as lay opinion testimony under MRE 701.
First, Gibson’s testimony was rationally based upon his perception. Gibson was not at the scene while the video footage was being recorded and did not observe first hand the events depicted on the video. Instead, Gibson perceived the video, produced short clips of the individuals while they were inside the store, and isolated certain frames to create still images. Based upon his scrutiny of the video surveillance footage and the still images he created from the video, Gibson provided his opinions regarding the identity of individuals within the video as compared to the still images from portions of the video. In Begay, the Ninth Circuit held that the contention that the officer’s testimony about the videotape was not based on his own perceptions because he was not present when the events that were videotaped occurred lacked merit because his testimony was based upon his own perceptions of the video itself. Begay, 42 F3d at 502-503. The Ninth Circuit particularly noted the officer’s “extensive review” of the video. Similarly, here, while Gibson was not at the scene when the events depicted in the video were occurring, Gibson testified he created the still photos from the surveillance video and cropped some of the photos to create a closer view. The purpose for creating the still photos was to determine whether the two suspect individuals had come to the BP gas station earlier in the evening before the murder took place. As was the conclusion with the officer in Begay, it can be inferred from Gibson’s testimony that he viewed the video and the still photos several times in order to draw his conclusions and opinions about the identity of the individuals in the surveillance video and still photos as compared to other individual depicted in the same evidence.
Second, Gibson’s testimony was intended to provide a clearer understanding about whether the two suspect individuals depicted in the video had been to the BP gas station earlier in the evening before the murder took place, a fact at issue in the case. In Begay, the Ninth Circuit concluded that the officer’s testimony likely helped the jury evaluate the videotape because it could reasonably be assumed that “one viewing of the videotape of a demonstration involving over 200 people” would not be sufficient for a jury to adequately discern the events depicted, and the officer’s testimony “could help the jury discern correctly and efficiently the events depicted in the videotape.” Begay, 42 F3d at 503. Similarly here, there were approximately six hours of surveillance video which Gibson reviewed to create still photographs and short clips. It is not clear whether there were other individuals in the six hours of video, but given its length, it can be inferred that there were. Because it can be inferred that Gibson viewed the surveillance footage and still photos several times to reach his conclusions and opinions, it can similarly be reasonably inferred that Gibson’s testimony helped the jury to correctly and efficiently determine whether the two individuals seen earlier in the footage were the same individuals who were involved in the murder later depicted in the video.
Third, Gibson’s testimony did not invade the province of the jury. In United States v. LaPierre, 998 F.2d. 1460, 1465 (CA 9th, 1993), an officer provided lay opinion testimony that the defendant was the individual captured in surveillance photographs from the bank that was robbed. The Ninth Circuit concluded that the trial court abused its discretion in admitting this testimony and remanded the case. The Ninth Circuit identified two situations under that circuit’s precedent illustrative of when such testimony was admissible. The LaPierre Court opined that the “common thread” of this authority was “reason to believe that the witness is more likely to identify correctly the person than is the jury.” The court concluded that the issue of whether the defendant in the courtroom is the person pictured in a surveillance photo “was a determination properly left to the jury.” See also United States v. Rodriguez-Adprno, 695 F3d 3240 (CA, 2012), holding that where a witness is in no better position than the jury to make an identification from a video or photograph, the testimony is inadmissible under FRE 701. But unlike the officer in LaPierre, Gibson did not identify defendant in the video or still images. Gibson’s testimony only linked individuals depicted in the surveillance video as being the same individuals depicted in the still photographs. While “a witness cannot express an opinion on the defendant’s guilt or innocence of the charged offense[,]” People v. Bragdon,142 Mich. App. 197, 199; 369 NW2d. 208 (1985), Gibson expressed no such opinion. Further, because Gibson was comparing the video surveillance video to still images that he himself had created from the six-hour long video, Gibson was in the best position to identify the individuals in the photographs as being the same as those depicted in the video. Gibson’s testimony did not invade the province of the jury.
Because Gibson’s testimony was (1) rationally based his own perception of the video and (2) helpful for the jury to determine whether the two individuals seen committing the crime in the surveillance video had come to the BP gas station earlier in the evening, Gibson’s testimony was admissible pursuant to MRE 701. Further, because we conclude Gibson’s testimony regarding his opinions and conclusions did not invade the province of the jury, we also conclude the trial court did not abuse its discretion in admitting Gibson’s testimony.
This case is interesting because while it approved of and allowed the image comparison evidence, the court ruled the evidence to be lay witness testimony rather than expert evidence. This case has been followed numerous times in Michigan [iv]. Most cases have found this evidence to be expert evidence. This does not mean that the Court of Appeals of Michigan devalued the comparison evidence but rather it approved of its entry via an alternative legal method.
In State of Ohio v. Stone, 2014 WL 5018831 (Ohio App. 1 Dist.), a multiple robbery case, the State called a forensic video analyst who gave expert evidence regarding a comparison she undertook involving the jacket worn by the robber as observed in the surveillance video and a jacket seized at the defendant’s home. The expert testified that as a result of her comparison process she determined that the jacket observed in the surveillance video was in fact the same jacket seized by the police. On appeal, the defendant argued that the analyst should not have been allowed to give expert opinion evidence. The Court of Appeals of Ohio did not specifically address the issue of the whether the trial judge erred in permitting the expert evidence as it found that there was sufficient other evidence that would have allowed the jury to render a decision on the identification issue in any event. Had the Court ruled on this issue, it is very unlikely that any error would have been found in the trial judge’s exercise of discretion as the expert was properly qualified and the comparison evidence was of assistance to the jury.
The Supreme Court of Minnesota, in State of Minnesota v. Ali, 855 N.W.2d 235 (2014, Supreme Court of Minnesota), reviewed a triple homicide case wherein the police gathered surveillance video from several establishments including the market where the homicides occurred. The State retained forensic video analysts from a private lab who used the surveillance videos to create a timeline for the movement of the suspects. Further, the video evidence was clarified and a comparison was undertaken as between questioned and known images, focusing on skin tone, overall build, clothing and footwear. At trial, the defendant argued that the expert forensic video analysis evidence should not be admitted based on the view that the trier of fact did not need expert assistance in order to properly assess the video evidence.
The trial judge ruled that because the experts clarified the images, it was important that they be able to testify about what clarifications were undertaken and whether any such clarifications affected the accuracy of the images. The trial judge further ruled that it was appropriate for the experts to testify regarding their comparisons and whether any those comparisons were impacted by the clarification processes. Further, if there were differences between images and those differences could be explained other than by saying they are in fact different people, objects, etc., the experts should be permitted to give such explanations.
As for whether the jury needed expert assistance, the trial judge ruled at page 250:
[T]he lighting, the aspect ratio, whatever goes into their opinion about why it might appear differently. That generally is not within the province of the normal juror as to how lighting and other factors on a video tape could affect the appearance of items from one video tape to the next.
The trial judge also placed limits on the experts’ testimony as follows at pages 250-251:
These witnesses may not … testify that they are in fact the same clothing or the same persons. They can point out similarities. They can say why they are similar and whether it was caused by their digital manipulation of the evidence or not. They may point out differences and explain why they appear to be different. They may not offer opinion that they are in fact the same clothing or same person’s [sic] depicted in the various video tapes.
On appeal, the Supreme Court reviewed the evidence given by the analysts in conjunction with the limitations imposed by the trial judge and ruled as follows at page 252:
Here, the district court concluded that the testimony would assist the jury because the videos reflected “similarities between clothing, build, skin tone, shoes,” and the experts had an “obligation to testify that those similarities are not caused by the digital clarification process.” Although some of these similarities are readily apparent from watching the surveillance videos, our review of the videos confirms that some of the finer details required expertise and were not within the knowledge of an average juror. For example, one of the Target analysts testified that by using the camera angle and distance from the camera, it is possible to tell that one suspect is taller than the other. The district court determined this sort of testimony would help the jury, and we hold that the court did not abuse its discretion in reaching that determination.
This decision is very helpful because it acknowledges the value that forensic video analysts can bring to the fact-finding function of the court and that if there are concerns regarding giving evidence on the ultimate issue, limitations can be imposed to address any such concerns.
A properly trained forensic video analyst knows the appropriate limitations when engaging in image comparison. When I teach for LEVA (Law Enforcement and Emergency Services Video Association International Inc.), I caution students against making positive identifications of people as such conclusions are better reached by the trier of fact after hearing the comparison evidence and the submissions of counsel at the end of the case. Conversely, forming positive conclusions on vehicles and objects are appropriate when the image comparison justifies such a result.
In Part 2 of “Image Comparison Evidence: The Canadian Experience”, I will discuss the Canadian experience and restrictions that have been placed on image comparison evidence in certain cases. I will also discuss how analysts can avoid having their evidence restricted in most cases and how to deal with restrictions when they arise.
i State of Iowa v. Piper, 2003 Iowa App. LEXIS 886 (2003, Iowa Court of Appeals)
ii United States v. Lightfoot, 2001 U.S. App. LEXIS 5121 (2001, United States Court of Appeals, Fourth Circuit); United States v. McKreith, 140 Fed.Appx.112 (2005, United States Court of Appeals, Eleventh Circuit)
iii United States v. Mosley, 1994 U.S. App. LEXIS 23969 (1994, United States Court of Appeals, Ninth Circuit); United States v. McKreith, 140 Fed.Appx.112 (2005, United States Court of Appeals, Eleventh Circuit); State of Texas v. Gonzales, 2006 WL 820387, not reported in S.W.3d (Court of Appeals of Texas, Fort Worth); State of Michigan v. Fomby, 2013 WL 1137075, 300 Mich.App. 46, 831 N.W.2d. 887 (Court of Appeals of Michigan)
iv Fomby was cited with approval in State of Michigan v. McLilly, 2015 WL302676 (Court of Appeals of Michigan) as well as State of Michigan v. Lucas, 2014 WL 4723823 (Court of Appeals of Michigan); State of Michigan v. Flake, 2014 WL 6602604 (Court of Appeals of Michigan); State of Michigan v. Harvey and Davis, 2015 WL 8953522 (Court of Appeals of Michigan)